README.md Raw History Folder

What is it?

Yet another coding agent harness, lightweight and written (vibe-slopped) in go.

one static binary.
built-in providers for Anthropic, OpenAI/Codex/Responses, Kimi, DeepSeek, Google Gemini/Vertex, GitHub Copilot, Bedrock, Azure OpenAI, OpenRouter, Groq, Cerebras, xAI, Together, Hugging Face, Mistral, Moonshot, Z.AI, Xiaomi, MiniMax, Fireworks, Vercel AI Gateway, OpenCode, Cloudflare AI, and Ollama/local models.
built-in tools (read, write, edit, bash, glob).
three run modes (interactive tui, print, json).
built-in telegram bot.
extensions in any language via subprocess + json-rpc. None installed by default; opt in with zot ext install or zot --ext. See docs/extensions.md.
user and extension themes via JSON; see docs/themes.md.
standing instructions via AGENTS.md files (global and per-project); see Persistent instructions.
reusable instructions via SKILL.md files; see docs/skills.md.
portable agents from local directories, .zot files, or temporary public GitHub downloads; see docs/zotfiles.md.

Install

One-liner (macOS, Linux)

curl -fsSL https://www.zot.sh/install.sh | bash

Detects your OS and architecture, downloads the latest release from GitHub, verifies the SHA-256 against the release's checksums.txt, extracts the binary, and drops it in /usr/local/bin, ~/.local/bin, or ~/bin, whichever is writable first. Pass a version or prefix to pin:

curl -fsSL https://www.zot.sh/install.sh | bash -s -- v0.0.1 ~/bin

One-liner (Windows, PowerShell)

iwr -useb https://www.zot.sh/install.ps1 | iex

Drops zot.exe into $HOME\bin and adds it to the user PATH if missing. Open a fresh terminal afterwards.

go install

go install github.com/patriceckhart/zot/cmd/zot@latest

The installed binary reports the tagged module version and supports zot update.

From source

git clone https://github.com/patriceckhart/zot
cd zot
make build        # produces ./bin/zot
make install      # into $GOPATH/bin

Prebuilt binaries

Every release on the releases page ships archives for Linux, macOS, and Windows on amd64 and arm64 (except windows/arm64), plus a checksums.txt file. Download, verify, chmod +x, and drop on your $PATH.

Authenticate

The easiest way is to just run zot and type /login. The TUI opens even without credentials and walks you through a browser-based login flow.

Credential lookup order

--api-key flag
provider-specific env var (ANTHROPIC_API_KEY, OPENAI_API_KEY, KIMI_API_KEY, MOONSHOT_API_KEY, DEEPSEEK_API_KEY, GEMINI_API_KEY, GOOGLE_API_KEY, GROQ_API_KEY, OPENROUTER_API_KEY, MISTRAL_API_KEY, XAI_API_KEY, CEREBRAS_API_KEY, TOGETHER_API_KEY, HF_TOKEN, ZAI_API_KEY, XIAOMI_API_KEY, MINIMAX_API_KEY, FIREWORKS_API_KEY, AI_GATEWAY_API_KEY, COPILOT_GITHUB_TOKEN, GITHUB_COPILOT_TOKEN, and others for provider-specific backends)
$ZOT_HOME/auth.json (API key or OAuth token; mode 0600)

$ZOT_HOME defaults to:

All platforms: $XDG_STATE_HOME/zot when XDG_STATE_HOME is set
macOS fallback: ~/Library/Application Support/zot
Linux fallback: ~/.local/state/zot
Windows fallback: %LOCALAPPDATA%\zot

API keys from commands

To keep an API key in a password manager instead of auth.json, configure an api_key_command for the provider:

{
  "anthropic": {
    "api_key_command": {
      "program": "op",
      "args": ["read", "op://Work/Anthropic/credential"],
      "timeout_ms": 120000
    }
  }
}

For a provider added through models.json, put the same credential object under additional_api_key_creds using its provider ID. program is executed directly, without a shell, so each argument must be a separate args entry. timeout_ms is optional and defaults to 120 seconds.

zot runs the command only when that provider is selected, not while checking login status or refreshing model catalogs in the background. Successful output is cached in memory for the rest of the zot process and is never written to disk. The command must print one non-empty line to stdout; zot removes trailing CR/LF characters, limits output to 64 KiB, and does not include command output in errors. Saving a normal key through /login replaces the command configuration, and /logout removes it.

Treat auth.json as executable configuration: anyone who can modify it can cause zot to run a program under your user account. zot does not interpret ! prefixes or execute command strings through a shell.

`/login` flow

Run zot and type /login. Pick one of two methods:

API key: a small local web server starts on 127.0.0.1:<free-port>, your browser opens a form, you pick a provider from the full API-key provider list, paste the key, and zot saves it to auth.json if accepted. Providers with a lightweight model-list endpoint are probed before saving; provider backends that need extra project/account env vars are saved directly.
Subscription: use your Claude Pro/Max, ChatGPT Plus/Pro, Kimi Code, SuperGrok/X Premium, or GitHub Copilot subscription. DeepSeek and Google Gemini do not have a subscription login path. For those, use the API-key flow.
- Anthropic and OpenAI pin the browser callback to fixed provider-specific ports (localhost:53692 for Anthropic, localhost:1455 for OpenAI) because those are the only ports their auth servers will redirect to.
- Anthropic uses the Claude Code OAuth flow. Messages go to api.anthropic.com with a bearer token and the Claude Code identity headers.
- OpenAI uses the Codex CLI OAuth flow. Messages go to chatgpt.com/backend-api/codex/responses with the chatgpt-account-id extracted from the returned id_token.
- Kimi uses the Kimi Code device-code OAuth flow. zot opens the verification URL, polls until you approve it in the browser, then sends messages to api.kimi.com/coding/v1 with the Kimi Code identity headers.
- xAI uses a device-code OAuth flow. zot opens a prefilled authorization URL, polls for approval, and uses the resulting token with the xAI API.
- GitHub Copilot uses GitHub's device-code login flow. zot stores the GitHub access token and exchanges it for short-lived Copilot inference tokens on demand.

Note on subscription login. The OAuth client IDs used are the ones published in Anthropic's Claude Code CLI, OpenAI's Codex CLI, Kimi Code CLI, xAI's device flow, and GitHub Copilot's device-code flow. Reusing them from a third-party tool may be against their terms of service and may be revoked at any time. Use it at your own risk; the API-key flow is the safe default.

Token refresh

OAuth access tokens are short-lived (Anthropic ~8h, OpenAI ~30d; Kimi, xAI, and GitHub Copilot also use refresh/exchange flows). zot refreshes or exchanges them automatically:

At every credential lookup, zot checks the stored expiry and, if past it (with a 60s safety margin), hits the provider's oauth/token endpoint with the stored refresh_token, persists the new access_token, refresh_token, and expiry back to auth.json, and hands the fresh token to the client.
The telegram bridge additionally refreshes once per turn so a bot that runs for days keeps working without manual intervention.
If the refresh itself fails (the refresh_token was revoked, or the account was logged out everywhere), the error bubbles up to the caller: the TUI shows it in the status line, the bot replies with it in your DM. Run /login to get a fresh token pair.

All data lives under $ZOT_HOME:

$ZOT_HOME/
├── config.json         # last-used provider/model/theme, saved automatically
├── auth.json           # api keys and oauth tokens (mode 0600)
├── sessions/           # jsonl transcripts, one dir per cwd
├── models-cache.json   # live /v1/models discovery cache (6h ttl)
├── AGENTS.md           # optional: global instructions appended to the prompt
├── SYSTEM.md           # optional: replaces the default system prompt
├── skills/             # optional: user SKILL.md files
├── themes/             # optional: user theme JSON files
├── extensions/         # installed extensions, one dir per extension
└── logs/               # app log files

Drop a SYSTEM.md in $ZOT_HOME to replace the built-in identity and zot-docs guidance for every run. --system-prompt still wins per-invocation. Pass an empty value (--system-prompt "") to intentionally omit the built-in identity. Custom prompts still receive appended instructions and generated context, including AGENTS.md, skills, auto-swarm guidance when enabled, and the date/cwd footer. Delete the file to revert to the default.

HTTP proxy

To route zot-managed HTTP and HTTPS requests through one proxy, add http_proxy to $ZOT_HOME/config.json:

{
  "http_proxy": "http://127.0.0.1:7890"
}

The setting is applied at startup to both HTTP and HTTPS traffic. Existing HTTP_PROXY, HTTPS_PROXY, http_proxy, and https_proxy environment variables take precedence for their corresponding protocol. NO_PROXY and no_proxy continue to control bypasses. Restart zot after changing the config file. If the URL contains proxy credentials, prefer protected environment variables because config.json is not a credential store.

Persistent instructions (AGENTS.md)

Use AGENTS.md to give zot standing instructions that layer on top of the default system prompt, without replacing it. This is the friendliest way to shape behavior (for example, taming local models that jump straight to code edits) because it adds guidance rather than taking over the base identity the way SYSTEM.md does.

zot discovers AGENTS.md files automatically at startup and loads them in this order:

$ZOT_HOME/AGENTS.md (global, machine-wide instructions that apply to every project).
Every AGENTS.md from the filesystem root down to the current working directory. More specific (deeper) files may override earlier ones. This includes ~/AGENTS.md when the working directory is inside your home directory.

All discovered files are appended to the prompt in that order, so a global baseline can be refined per project. To list the active Zotfile agent, loaded context paths, extensions, and user-installed skills above the interactive transcript, enable show loaded resources at startup in /settings. The agent section appears only for zot run, the setting is off by default, built-in skills are omitted, and the displayed lists are not sent to the model or stored in the session transcript.

For a general, non-project-specific instruction set, put your rules in $ZOT_HOME/AGENTS.md, for example:

Treat questions and discussions as requests for explanation. Do not edit files or run tools unless explicitly asked to make a change. Ask before modifying code.

AGENTS.md vs the other mechanisms:

Mechanism	Scope	Effect
`$ZOT_HOME/AGENTS.md`	global, every run	appended to the default prompt
`./AGENTS.md` (and parent dirs)	project	appended to the default prompt
`$ZOT_HOME/SYSTEM.md`	global, every run	replaces the built-in identity and zot-docs guidance
`--append-system-prompt <text>`	single run	appends for one invocation (repeatable)
`--system-prompt <text>`	single run	replaces the built-in identity and zot-docs guidance

Note: zot does not read a CLAUDE.md instruction file. The only Claude-compatible thing it picks up is skills under .claude/skills/. If you are migrating from Claude Code, move that content into AGENTS.md (global or per-project) and zot will use it.

Changelog on update

The first time you launch a newer zot binary, the TUI shows the GitHub release notes once in a dismissible overlay. Press any key to close. The version is recorded in config.json's last_changelog_shown so the same release notes never reappear. Fresh installs don't see a changelog (no upgrade has happened yet). The fetch is best-effort: a network failure or a missing release page silently skips, with another attempt on the next launch.

Usage

zot                              # interactive tui
zot "fix the failing test"       # tui, pre-filled prompt
zot -p "list all go files"       # print final text, exit
echo "list all go files" | zot   # piped stdin implies print mode
cat README.md | zot -p "summarize this text" # combine stdin with the prompt
zot -p --stats stats.json "task" # print final text and write generation stats
zot --json "refactor main.go"    # newline-delimited json events, exit
zot --continue                   # resume the most recent session for this cwd
zot --resume                     # pick a session to resume
zot --list-models                # show supported models
zot --help

Print-mode stats contain provider, model, prompt_tokens, reasoning_tokens, generated_output_tokens, and elapsed_ms. Counts cover all model turns triggered by the prompt, including tool loops. Prompt tokens include cache reads and writes. reasoning_tokens is null when the provider does not report a separate count; in that case generated_output_tokens is the provider's total output count and may include reasoning. Elapsed time covers the agent run, not startup and credential resolution. The file is written only after a successful run.

Flags

Flag	Description
`--provider <id>`	Pick the provider (for example `anthropic`, `openai`, `openai-codex`, `kimi`, `google`, `github-copilot`, `groq`, `openrouter`, `amazon-bedrock`, `ollama`; see Providers).
`--model <id>`	Pick the model (see `--list-models`).
`--api-key <key>`	Override the API key.
`--base-url <url>`	Override the provider base URL (tests, self-hosted).
`--insecure`	Skip TLS certificate verification for the explicit `--base-url` endpoint or a `baseUrl` defined for a user model in `models.json` (self-signed local/internal inference servers). Built-in providers, auth, and model discovery keep normal TLS verification.
`--system-prompt <text>`	Replace the default system prompt for this run (also overrides `$ZOT_HOME/SYSTEM.md`; pass `""` for no built-in identity).
`--append-system-prompt <text>`	Append text to the system prompt (repeatable).
`--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max`	Set the reasoning level on supported models (default: off). `max` is a separate opt-in tier above `xhigh`.
`--stats <path>`	With `-p`/`--print`, write generation statistics as JSON.
`-c`, `--continue`	Resume the latest session for this cwd.
`-r`, `--resume`	Pick a session to resume.
`--session <path>`	Resume a specific session file.
`--no-session`	Don't read or write session files.
`--cwd <path>`	Use `<path>` as the working directory.
`--no-tools`	Disable all tools.
`--tools <csv>`	Only enable the listed tools.
`--max-steps <n>`	Cap agent loop iterations (default 50).
`-e`, `--ext <path>`	Load an extension from `<path>` for this run (repeatable; wins against installed extensions of the same name).
`--no-ext`	Skip extension discovery for this run. `--ext` still works on top, so `--no-ext --ext ./x` runs only `x`.
`--no-skill`	Disable all skills, including built-ins. No `skill` tool is registered and the system prompt has no skill manifest.
`--no-context-files`, `-nc`	Disable discovery and loading of global and project `AGENTS.md` files.
`--no-yolo`	Confirm every tool call before it runs (interactive TUI only). A dialog shows the tool name and a one-line preview of its args, while edit calls show the proposed diff in the tool panel, with four choices: yes, yes-always-this-tool-this-session, yes-always-this-session, no. Press `/` while the dialog is focused to run a slash command first; closing its input or child dialog returns to the pending confirmation. Ignored with a stderr warning in print / json / rpc modes, where tools still run freely so scripts and automation keep working.

Tools

read: read text files, or inline images (PNG, JPEG, GIF, WebP).
write: create or overwrite files, making parent directories as needed.
edit: one or more exact-match replacements in an existing file.
bash: run a command in the session cwd with merged stdout/stderr and a timeout. On Unix, zot uses /bin/bash -c when available, then bash -c from PATH, and falls back to POSIX /bin/sh -c when Bash is unavailable. On Windows, it uses cmd /C. macOS ships Bash 3.2 by default, so newer Bash features may be unavailable.

When the sandbox is on (see /jail), all four tools refuse paths outside the session cwd.

Modes

Interactive (default): chat TUI with streaming output, spinner, cost meter, slash commands.
Print: zot -p "prompt" runs the agent to completion and writes only the final assistant text to stdout.
Stream: zot --stream "prompt" runs without the TUI and writes assistant text to stdout as it arrives. Tool activity goes to stderr.
Piped input: when no mode is specified, piped stdin selects print mode. Print, stream, and JSON modes prepend piped stdin to the positional prompt, separated by a newline. For example, echo "list all go files" | zot or cat README.md | zot -p "summarize this text".
JSON: zot --json "prompt" emits one JSON object per agent event to stdout, newline-delimited. The schema is documented in docs/rpc.md.
RPC: zot rpc runs as a long-lived child process; commands in on stdin, events and responses out on stdout, both as NDJSON. Designed for embedding zot in third-party apps written in any language. See docs/rpc.md for the wire schema and examples/rpc/{python,node,shell,go} for working clients.

zotfile agents

A zotfile packages an agent's instructions, skills, requirements, and enforced tool permissions as a shareable agent. Run one from a local directory, a packed .zot artifact, a short name, or directly from a public GitHub repository:

zot run ./my-agent
zot run ./my-agent.zot
zot run frkr/zot-archify
zot run acme/agents/code-reviewer --cwd /path/to/project

Single-part names resolve only to a matching local directory or .zot archive. Anyone can publish an agent as a public GitHub repository and run it as owner/repository, or publish a collection and run an agent directory as owner/repository/agent. zot has no built-in owner, official collection, configured registry, or allowlist. For GitHub sources, zot downloads the repository archive into a temporary directory, validates and runs the selected agent, then removes the downloaded source when the command exits. Agent data, consent receipts, and sessions still persist under $ZOT_HOME. See docs/zotfiles.md for authoring, permissions, packaging, and current limitations.

Embedding

Two ways to drive zot from another program:

Go in-process: import github.com/patriceckhart/zot/packages/agent/sdk. One Runtime per project; Prompt(ctx, text, images) returns a channel of Event. Small example in examples/sdk/.
Any language, out-of-process: spawn zot rpc as a subprocess and exchange newline-delimited JSON over its stdin/stdout. Wire format and event schema in docs/rpc.md. Reference clients live under examples/rpc/.

Both interfaces share the same event schema, so transcripts captured by one can be replayed through the other.

Slash commands

Slash command names are case-insensitive in the TUI and messaging backends; arguments keep their original case. Type / in the TUI to open the autocomplete popup. Available commands:

Command	Description
`/help`	Show key bindings and commands.
`/login`	Log in via API key or subscription (opens a dialog).
`/logout [provider]`	Clear credentials for any logged-in provider, or all when omitted. `/logout openai-codex` clears ChatGPT/Codex subscription auth while preserving a public OpenAI API key; `/logout kimi` also disables fallback to the official Kimi Code CLI token until you log in to Kimi through zot again.
`/model`	Pick a model from a list (or `/model <id>` to set directly).
`/reasoning`	Set the reasoning level for subsequent model calls.
`/llama`	Connect to the configured llama.cpp router, load, unload, or remove cached models, and search/download GGUF models from Hugging Face with live progress. Shown after llama.cpp login is configured.
`/sessions`	Resume a previous session for this directory.
`/session`	Five ops on the current session: inspect its `timeline`, `export` to a portable `.zotsession` file, `import` one back in, `fork` from a past user message, or view its branch `tree`. Opens a picker without an argument; direct forms include `/session timeline`, `/session export [path]`, `/session import <path>`, `/session fork`, and `/session tree`. Default export destination is `~/Downloads`.
`/jump`	Scroll the chat to a previous turn (or `/jump <text>` to filter).
`/btw`	Side chat with full context that doesn't add to the main thread.
`/swarm`	Spawn, monitor, and chat with background subagents. Each runs in parallel with your main session and shares its working directory.
`/skills`	List discovered skills (SKILL.md files) and preview their bodies.
`/compact`	Summarize the transcript into one message to free up context.
`/study`	Run the canned prompt "Read and understand everything in the current directory." so the agent has full project context before you start asking targeted questions. Pass a path — typed, drag-dropped, or selected via `@` — to target a specific file or directory instead: `/study [dir:packages/]`, `/study cmd/zot/main.go`.
`/jail`	Confine tools to the current directory.
`/unjail`	Allow tools to touch paths outside again.
`/reload-ext`	Hot-reload all extensions (re-read manifests, respawn subprocesses, rebuild tool registry).
`/telegram`	Connect, disconnect, or show status of the Telegram bridge (takes `connect` / `disconnect` / `status` as an optional argument; opens a picker without one). When connected, DMs from the paired user become prompts in the running session and the assistant's replies are mirrored back to Telegram. Alias: `/tg`.
`/settings`	Change persistent settings, including inline images, auto-swarm, and the auto-compact threshold. Saved to `$ZOT_HOME/config.json`; takes effect immediately.
`/clear`	Clear the chat transcript.
`/exit`	Exit zot.

Extension-registered commands appear under a divider at the bottom of the popup, sorted by name.

Shell escape (`!command`)

Type ! followed by a command to run it directly without going through the model. Everything after the ! is passed to the same shell the bash tool uses (/bin/bash -c when available on Unix, then bash -c from PATH, with POSIX /bin/sh -c as a fallback; cmd /C on Windows), runs in the session working directory, and honors the /jail sandbox. The command, merged output, and exit status are appended to the transcript as user context, so the model can use them on the next turn. Running the command does not itself start a model turn. A running !command shares the busy state with the agent: esc cancels it, and you cannot start one while a turn (or another shell escape) is in flight.

`/sessions`

Shows previous sessions for the current working directory, newest first, with timestamp, model, message count, cost, and the first user prompt. Pick one with up/down, enter to resume, esc to cancel. zot swaps the current session file for the selected one and replays the full transcript (including tool calls) into the agent. Sessions remember the model they ended on, so resuming picks up on that exact model even if your global default changed.

`/session`

Five ops on the current session. /session alone opens a picker; each is also runnable directly.

/session timeline. Replaces the chat area with a read-only timeline of the current system prompt, user and assistant messages, and tool calls. The header shows provider-reported context usage and an estimated split between system prompt, tool definitions, and messages. The split uses a simple byte estimate, so it is directional rather than tokenizer-exact. Use up / down or pgup / pgdn to select an event, tab / shift+tab to inspect its summary, payload, result, schema, and transcript-derived timing, and / to search. ctrl+e writes a JSON export to ~/Downloads (falling back to the home directory) with 0600 permissions on platforms that support Unix file modes. Images are represented by MIME type and byte size instead of embedding their data. Press esc to return to chat.
/session export [path]. Writes the running transcript to a portable .zotsession file. Default destination is ~/Downloads/<timestamp>-<session-id>-<prompt-slug>.zotsession. Pass a path to override; a directory is fine (a dated name is built inside), a bare name gets .zotsession appended. The meta's cwd is stripped on the way out so the recipient doesn't see your filesystem layout.

What's included. Only the main chat thread of the running session — messages, tool calls, tool results, compactions, and usage. /swarm subagents are NOT included. Their transcripts, unix-socket inboxes, and per-agent session files are all machine-local; a .zotsession is just a chat transcript and has no way to revive a unix socket on another box. If you want the conversation, copy it out of the dashboard manually.
/session import <path>. Copies a .zotsession file into $ZOT_HOME/sessions/<cwd-hash>/ with a fresh id and the current cwd, then switches the running agent onto it. Imported sessions are first-class: they show up in /sessions, /jump, and the tree. Drag-drop paths in the editor are accepted (zot strips the surrounding quotes automatically).
/session fork. Opens a turn picker (same shape as /jump). Pick any past user message; zot copies every message up to and including that turn into a new session, records parent + fork_point in the new meta, and switches onto the branch. The parent session stays on disk. Use it to try a different question without polluting the original transcript, or to rewind after the agent went down the wrong path.
/session tree. Shows every session in the current cwd arranged by parent/child relationships, depth-first with indent per level. The current session is tagged [current]. Pick any entry to switch into it. Parentless sessions are roots; branches created via /session fork nest under whichever session they were forked from. Orphaned children (whose parent file was deleted) still show as roots so they stay discoverable.

`/jump`

Opens a turn picker for the current session, one row per user prompt, each showing the turn number, how many tools that turn invoked, and the first line of the prompt. up/down to pick, enter to jump, esc to cancel. Any printable rune while the picker is open extends a filter; backspace narrows it back. /jump <text> pre-applies the filter; if exactly one turn matches, zot jumps straight there without showing the picker.

Jumping is non-destructive. The transcript is untouched, the viewport just scrolls so the chosen turn is at the top. A muted line at the top of the chat reads viewing turn N of M, pgdn to catch up. Scroll back to the bottom with pgdn (or keep scrolling with the arrow keys) and the indicator goes away.

`/btw`

Opens a side-chat overlay with the full main session as frozen context, so you can ask quick clarifying questions ("does asyncio.gather() catch exceptions?", "btw the bundle budget is 10MB", "what's the default fetch timeout?") without bloating the main thread.

Each question runs an isolated agent turn against system + main transcript + side-chat history so far. The side chat has the same tools and guards as the main chat, including per-tool confirmation when --no-yolo is active. Responses, tool calls, and tool results stay in the overlay. When you press esc to close, nothing has been added to the main session and subsequent main-thread turns don't re-read any of the side-chat exchanges, keeping the running context window lean.

/btw                              # open the overlay, type questions interactively
/btw does PUT replace the whole resource?

Inside the overlay: enter sends, esc cancels an in-flight call (or closes the overlay if idle), ctrl+c closes immediately. Side-chat exchanges never touch the transcript and aren't persisted to the session file.

`/swarm`

Background subagents that run alongside your main session. Each one is a separate zot subprocess with its own model loop, its own persistent session file, and its own chat in the dashboard — but they all run in the same working directory as the host, so they see and edit the same files you do. Spawn one for a side task (“draft the migration”, “investigate this stack trace”, “write the test harness for module X”), keep going in the main thread, check in on it whenever you want.

Agents edit the same files you do. They use the same read / write / edit / bash tools as the main agent against the host's working directory. There's no per-agent worktree or branch. If you need parallel edits on isolated checkouts, set that up yourself with git worktree outside zot.

/swarm                            # open the dashboard
/swarm new <task>                 # spawn an agent
/swarm new --model gpt-5 <task>    # pin the new agent to a specific model
/swarm logs <id>                  # jump straight into one agent's transcript
/swarm send <id> <text>           # send a follow-up without opening the dashboard
/swarm resume                     # pick a stopped agent to bring back
/swarm resume <id>                # bring a specific agent back
/swarm kill <id>                  # stop a running agent (its state stays)
/swarm remove <id>                # delete the agent's session and state
/swarm list                       # alias for opening the dashboard

Dashboard (/swarm with no arg) — a list of every agent for the current session, with status, age, and current activity. Keys:

Key	Action
`↑` / `↓`	Move cursor between rows.
`enter`	Open the highlighted agent's transcript view.
`n`	Spawn a new agent (opens an inline task editor; inherits the host's current model).
`p`	One-off prompt editor for the selected row (without entering the transcript).
`R`	Resume a stopped agent in place.
`k`	Kill the selected running agent. Its session and state stay so you can resume it later.
`r`	Remove the selected agent entirely (session + meta gone).
`esc`	Close the dashboard.

Inside an agent's transcript — a chat overlay with an always-on inline composer at the bottom. The conversation flows above it; type and enter to send a follow-up. The view auto-follows streaming output and shows an inline spinner with the agent's current activity (thinking, tool: edit_file, etc.) while it's busy. esc returns to the dashboard.

Switching the spawn model from inside the editor — while composing a task in the n-prompt, type /model on its own line and enter. The standard /model picker pops up; pick a model, the picker closes, and the editor reopens with your typed task intact and the new model pinned for the spawn.

Session scoping — each agent is stamped with the host session that spawned it and only shows up in that session's dashboard. Swap sessions with /sessions and the dashboard re-narrows accordingly. Agents from other sessions keep running in the background and reappear when you switch back.

Persistence across zot restarts — every spawn writes a meta.json next to its event log and session file under $ZOT_HOME/swarm/agents/<id>/. On the next zot launch they show up in the dashboard as detached; press R (or /swarm resume <id>) to bring one back. Resumed agents reattach to the same session and inbox socket, so the conversation continues from where it left off.

Where state lives — everything per-agent (session file, events log, inbox socket, meta) lives under $ZOT_HOME/swarm/agents/<id>/. The agent's actual code edits land directly in your repo; track them with normal git status / git diff.

/session export does NOT bundle subagents. A .zotsession is just the main chat transcript; per-agent state (session file, unix-socket inbox) is machine-local and doesn't round-trip through a JSONL file. To share what an agent said, copy it out of the transcript view manually.

Auto-swarm. With /settings -> auto-swarm on, the main agent gets a built-in swarm_spawn tool and a system-prompt nudge to use it. It can then fork sub-agents on its own when a request naturally splits into independent parallel work ("implement A and B", "investigate three files"). Each spawn returns the sub-agent id immediately and the main turn keeps going. When every sub-agent the agent spawned in that batch finishes its initial task, zot injects one [auto-swarm update] message back into the main chat recapping each agent's status, task, and complete final response. If an agent produced no assistant response, zot includes a truncated transcript tail for diagnostics instead. The main agent then writes a short follow-up summary referencing the agents by id. Off by default; toggle from /settings.

`/settings`

Opens a dialog with every persistent setting. up/down to navigate, enter or space to change the selected row, esc to close (rows that open a sub-view, like model shortcuts, use esc to go back one level first). Changes are written to $ZOT_HOME/config.json and take effect on the next turn (no restart needed). Current settings:

render images when supported — draw screenshots / read-returned images inline using the terminal's image protocol, or fall back to a text placeholder. Auto-detected from TERM_PROGRAM; the toggle overrides the detection. The row is greyed out and forced off on terminals that don't speak any image protocol.
auto-swarm — let the main agent spawn background sub-agents in parallel via a built-in swarm_spawn tool. Off by default. When on, the tool is registered with the running agent, the system prompt gains a short addendum telling the model to delegate independent sub-tasks proactively, and zot watches every sub-agent the main agent spawns. As soon as the last sub-agent in a batch finishes its initial task, an [auto-swarm update] message is injected back into the chat with each agent's status / task / transcript tail, so the main agent can summarise the collective outcome. Flipping off mid-session removes the tool from the live agent and strips the addendum on the next turn — the model stops trying to delegate. See /swarm for the dashboard that lets you monitor, message, kill, or remove the spawned agents.
auto-compact threshold — choose off, 70%, 80%, 85% (default), or 90% of the model's advertised context window. The selected percentage controls automatic compaction before and after interactive turns and persists as auto_compact_threshold. off disables percentage-based triggers but keeps manual /compact and automatic recovery from context-window and payload-too-large responses.
jail new sessions by default: start every new agent with tools confined to its working directory. Off by default. The setting applies to interactive, print, JSON, RPC, and background-agent runs, persists as jail_by_default, and immediately updates the current interactive session. /jail and /unjail remain session-scoped overrides and do not change this default.
OpenRouter Server Tools: opt in to OpenRouter's beta web search, web fetch, advisor, subagent, fusion, image generation, datetime, and model-search tools. Off by default and available only with the OpenRouter provider. zot limits the remote loop to four server-tool calls per turn. --no-tools always disables them. Presets keep their own tool configuration and are not overridden by this setting. Server tools run remotely without zot's local tool confirmation dialog and may incur additional OpenRouter charges.
compact transcript rendering: reduce visual chrome in the chat transcript. Tool calls render as a quiet header plus indented output instead of a bordered box, and sent messages render without padded background bubbles. Off by default. Changes apply immediately and persist to config.json as compact_mode.
show loaded resources at startup: list the active Zotfile agent (for zot run), loaded AGENTS.md paths, extensions, and user-installed skills in compact sections above the transcript. Built-in skills are omitted. Off by default. Changes apply immediately and persist to config.json as show_instructions_at_startup.
TUI settings: opens a sub-view for input layout and status placement. Input style can be plain (default prompt line), lines (separator lines above and below the input), or block (a user-bubble-style input block). Status position places model, usage, and working-directory information above or below the input. Working spinner position places the busy spinner above or below the input. Changes apply immediately and persist to config.json as tui_input_style, tui_status_position, and tui_working_position (above_input or below_input for the position fields).
reasoning level: choose reasoning for supported models: off, minimum, low, medium, high, xhigh, or max. The max tier is opt-in and sent natively to GPT-5.6 and adaptive-thinking Claude models; unsupported backends clamp it to their highest accepted effort. The change is persisted to config.json and applied to the next model call. Use /reasoning to open this selector directly. The selector only shows distinct levels supported by the active model; models without reasoning support only offer off.
color theme — choose the built-in auto/dark/light theme or any JSON theme discovered under $ZOT_HOME/themes or a loaded extension. Theme files can override any subset of UI colors, syntax colors, and spinner frames/messages. Changes apply immediately; if a selected theme file is deleted, zot resets to auto. See docs/themes.md.
model shortcuts — opens a sub-view with nine slots (model 1 ... model 9). enter on a slot opens the same /model selector and binds the chosen provider/model to that slot; backspace clears a slot. Once assigned, press Ctrl+1 ... Ctrl+9 from the editor to switch the active model instantly (the same cross-provider swap /model performs, transcript and cost carried over). Assigning a shortcut does not change the current model. Shortcuts are skipped while a turn is running.

`/skills`

Opens a picker listing every discovered SKILL.md file, built-ins hidden. Each row shows the skill name, source, and description. enter opens the body inline (scrollable with up/down/pgup/pgdn); esc goes back. Re-runs discovery each time it opens, so edits to a SKILL.md during a session are reflected immediately.

`/compact`

Sends the current transcript through the model with a structured summarization prompt. The returned summary replaces the transcript as one synthetic user message, with the last few exchanges kept verbatim for continuity. The status bar's context meter resets. Use it when the context meter creeps past ~80%.

zot also auto-compacts in the background: after any turn that reaches the configured context threshold, the agent kicks off a condense pass on its own. Choose off, 70%, 80%, 85% (default), or 90% under /settings → auto-compact threshold. You'll see condensing history, esc to cancel above the status bar and an (auto) tag next to the context percentage; esc aborts it without touching the transcript. Turning the percentage trigger off does not disable automatic compaction and retry after a context-window or payload-too-large response.

`/jail`

Enforces a sandbox rooted at the cwd shown in the status bar. read, write, and edit resolve their target path (including through symlinks) and refuse anything outside the sandbox. bash refuses obvious escape patterns (sudo, rm -rf /, leading cd /, cd .., cd ~, chmod -R, dd of=/, and similar) and rejects shell arguments or redirections that point outside the sandbox. The status bar shows jailed, ~/your/cwd while active. Enable jail new sessions by default in /settings to persist this behavior across launches; /unjail then unlocks only the current session.

This is a guardrail against accidents, not a hard security boundary. If you need real isolation, run zot under docker or a proper sandbox.

Sessions

Every interactive or print/json run (unless --no-session) writes a JSONL transcript under $ZOT_HOME/sessions/<cwd-hash>/. Resume any of them with --continue, --resume, --session <path>, or interactively via /sessions inside the TUI. Empty sessions (the user exited without prompting) are deleted on close so the list stays tidy.

Use zot sessions prune outside the TUI to find sessions whose recorded working directories no longer exist. The command groups sessions by directory, shows the number and human-readable total size of stored sessions, lets you select groups, and requires confirmation before permanently deleting files. It preserves sessions when a directory check fails with an error other than "not found" and rechecks each selected directory immediately before deletion. A deleted directory and a path hidden by some unmounted filesystems are indistinguishable, so review the selection and ensure remote filesystems are mounted before deleting missing-directory groups.

Use --older-than to prune sessions by last activity instead, including sessions for directories that still exist. Supported age units are m (minutes), h (hours), d (24-hour days), w (7-day weeks), mo (calendar months), and y (calendar years). Add --cwd PATH to limit age pruning to one working directory. The session file is checked again immediately before deletion and preserved if it has become active recently.

zot sessions prune --dry-run                         # list missing-directory groups
zot sessions prune --older-than 4h --dry-run         # inspect sessions inactive for four hours
zot sessions prune --older-than 1mo --cwd ~/project  # select old sessions for one directory
zot sessions prune --older-than 1y --all              # select every match, then confirm
zot sessions prune --older-than 30d --all --yes       # delete every match without prompting

--all selects every match and still asks for confirmation. Adding --yes makes deletion non-interactive and requires --all. Malformed, unreadable, symlinked, and non-absolute session entries are reported and preserved during missing-directory pruning. The scan includes normal sessions and named-agent sessions below $ZOT_HOME/sessions/; explicit session files stored elsewhere and swarm-agent state are not included.

Providers

zot's built-in provider catalog includes:

Subscription-capable: Anthropic Claude Pro/Max (anthropic), OpenAI Codex / ChatGPT Plus/Pro (openai-codex), Kimi Code (kimi), SuperGrok/X Premium (xai), GitHub Copilot (github-copilot).
Direct API providers: Anthropic, OpenAI Chat Completions, OpenAI Responses, DeepSeek, Google Gemini, Kimi/Moonshot, Moonshot CN, Groq, Cerebras, xAI, Together AI, Hugging Face Router, OpenRouter, Mistral, Z.AI, Xiaomi/MiMo token-plan regions, MiniMax global/CN, Fireworks, Vercel AI Gateway, OpenCode/OpenCode Go.
Cloud/platform providers: Amazon Bedrock, Google Vertex AI, Azure OpenAI, Cloudflare Workers AI, Cloudflare AI Gateway.
Local/compatible: Ollama, llama.cpp router mode, and OpenAI-compatible local endpoints via --base-url.

Use /login to store API keys or subscription credentials. /model only shows models from providers that are currently available from env vars, auth.json, Kimi CLI fallback, local Ollama, or a configured llama.cpp router.

Models

--list-models or the /model picker shows the full catalog across all built-in providers. Three sources:

Catalog: models baked into zot, covering Claude, GPT/Codex, Gemini/Gemma, Kimi/Moonshot, DeepSeek, Groq-hosted Llama/Gemma/Compound, OpenRouter-routed models, Bedrock model ids, Vertex model ids, Azure OpenAI deployments, Copilot models, and other provider-specific catalog entries.
Live: IDs discovered from GET /v1/models using your stored API key (cached for 6h in $ZOT_HOME/models-cache.json, refreshed in the background on startup).
Speculative: IDs that appear in the upstream generator but aren't live on the public API yet. They'll 404 today and start working the moment the provider ships them.

The context meter in the status line uses the model's advertised context window to show how much of it your last turn consumed.

Model fallback (rescue)

When a turn fails because of a recoverable provider error — expired token (401), permission denied (403), rate limit (429), provider outage (502/503/504), or a transient network failure — zot opens a rescue picker over the chat instead of just painting a red banner.

The picker is the same vertical list / fuzzy filter UI as /model, but it only shows models from providers you're currently logged in to (env vars, auth.json, Kimi CLI fallback, ollama). The failed model is excluded. Press ↑/↓ to choose, enter to retry the same prompt on the new model, esc to dismiss.

Before the actual provider request fires, the OpenAI / Anthropic / Kimi / DeepSeek / Google / OpenAI-Codex clients also do up to two silent retries with short backoff (250ms, 750ms) on 502/503/504 and connection-reset / EOF-before-headers errors. Most edge-proxy blips disappear without you ever seeing the rescue picker.

A rescue retry always drops launch-time --api-key and --base-url before rebuilding the agent. Those overrides are usually the reason the rescue triggered (bad key, typo'd base URL, corporate gateway only valid for the originally-picked provider), so the retry re-resolves credentials from env vars / auth.json / provider defaults instead. Use /model if you want overrides to stick.

No configuration is required — the candidate list is built dynamically from your active credentials. Bad-request / context-length / serialization errors are NOT routed to the rescue picker, because switching models won't fix them; those still surface as a normal error.

Custom models

Place a models.json in $ZOT_HOME ($XDG_STATE_HOME/zot/ when set, otherwise the platform default above) to add models that aren't in the baked-in catalog or to override existing entries:

{
  "providers": {
    "openai": {
      "models": [
        {
          "id": "gpt-5.5",
          "name": "GPT-5.5",
          "reasoning": true,
          "reasoningLevelMap": {"minimum": "low", "max": ""},
          "contextWindow": 400000,
          "maxTokens": 128000
        }
      ]
    }
  }
}

Supported fields per model: id (required), name, reasoning, reasoningLevelMap, contextWindow, maxTokens, baseUrl, priceInput, priceOutput, priceCacheRead, priceCacheWrite.

reasoningLevelMap is optional. Protocol defaults apply when it is omitted. Add only model-specific exceptions using minimum, low, medium, high, xhigh, or max as keys. Map a key to another level when both inputs are equivalent, use an identity mapping such as "max": "max" to enable a level beyond the protocol default, or map it to an empty string or off to remove it. The same effective mapping drives /reasoning and provider requests.

Provider keys are normalized: openai-codex and openai-responses map to openai, anthropic-messages maps to anthropic, moonshot, moonshot-ai, and kimi-code map to kimi, and deepseek-chat and deepseek-ai map to deepseek. Built-in provider ids such as groq, openrouter, github-copilot, amazon-bedrock, google-vertex, azure-openai-responses, fireworks, vercel-ai-gateway, mistral, and xai can also be used directly.

User-defined models show source: user in --list-models and take precedence over both the baked-in catalog and live-discovered models. Adding a models.json does not hide the built-in catalog; entries are merged on top of it. Missing or invalid files are silently ignored.

Custom providers

A top-level provider key that is not a built-in id defines a custom provider. Give it a provider-level baseUrl and an api wire format (openai for OpenAI-compatible Chat Completions, the default, openai-responses for the OpenAI Responses API, or anthropic for the Anthropic Messages API). A model-level baseUrl overrides the provider-level one for that model; an unknown api value falls back to openai with a warning.

{
  "providers": {
    "my-company": {
      "baseUrl": "https://llm.mycompany.com/v1",
      "api": "openai",
      "models": [
        { "id": "company-llm-v2", "name": "Company LLM v2" }
      ]
    }
  }
}

Custom providers are first-class: they appear in --list-models, /model, and /login. models.json never stores secrets. Supply the key through /login, --api-key, or a derived environment variable in upper snake case (so my-company reads MY_COMPANY_API_KEY). Because many self-hosted gateways do not expose a model-list endpoint, custom provider keys are accepted and stored without a verification probe; an invalid key surfaces on the first model call.

To retrieve this custom provider's key from a password manager, add a matching entry to $ZOT_HOME/auth.json:

{
  "additional_api_key_creds": {
    "my-company": {
      "api_key_command": {
        "program": "op",
        "args": ["read", "op://Work/OpenAI/credential"],
        "timeout_ms": 120000
      }
    }
  }
}

The provider IDs in models.json and auth.json must match. Then select the custom model directly:

zot --provider my-company --model company-llm-v2

Kimi Code

zot has built-in Kimi support through the Kimi Coding endpoint and Moonshot's OpenAI-compatible chat API.

zot --provider kimi

By default this uses:

model: kimi-for-coding
base URL: https://api.kimi.com/coding/v1

Credential lookup order for Kimi:

--api-key
KIMI_API_KEY
MOONSHOT_API_KEY
$ZOT_HOME/auth.json
the official Kimi Code CLI token at ~/.kimi/credentials/kimi-code.json, unless disabled by /logout kimi

Use /login for either API-key login or Kimi Code subscription login. The subscription flow uses Kimi Code's device-code OAuth flow: zot opens the verification URL, waits for browser approval, stores the token in auth.json, and refreshes it automatically.

For direct Moonshot API keys or a custom compatible endpoint:

zot --provider kimi --model kimi-k2-0905-preview --base-url https://api.moonshot.ai/v1 --api-key "$KIMI_API_KEY"

Kimi K3 is built in as kimi/k3, moonshotai/kimi-k3, moonshotai-cn/kimi-k3, opencode-go/kimi-k3, openrouter/moonshotai/kimi-k3, and vercel-ai-gateway/moonshotai/kimi-k3. Its output limit is 131,072 tokens on every built-in route.

You can add additional Kimi/Moonshot model IDs to models.json under the kimi provider.

xAI

xAI supports either XAI_API_KEY or /login subscription authentication. The subscription option is labeled Sign in with SuperGrok or X Premium, opens a prefilled device-authorization URL, and refreshes the stored token automatically. The default xAI model is grok-4.5, sent through the Responses API.

DeepSeek

zot has built-in DeepSeek support through DeepSeek's OpenAI-compatible chat API.

zot --provider deepseek

By default this uses:

model: deepseek-v4-pro
base URL: https://api.deepseek.com/v1

Catalog ships with deepseek-v4-pro (reasoning) and deepseek-v4-flash. These are exactly the IDs returned by GET https://api.deepseek.com/models today. You can add additional model IDs to models.json under the deepseek provider.

Credential lookup order for DeepSeek:

--api-key
DEEPSEEK_API_KEY
$ZOT_HOME/auth.json

Use /login and pick api key to paste a DeepSeek key. zot probes /v1/models once and stores the key under deepseek in auth.json.

Auth model: API key only. DeepSeek does not offer a subscription OAuth flow. The /login subscription step lists only Anthropic, OpenAI, and Kimi; DeepSeek shows up only under /login → api key.

Text only at the wire level. DeepSeek's chat-completions endpoint currently rejects the multimodal content schema (unknown variant image_url, expected text). When the active provider is deepseek, zot silently drops ImageBlock parts from outgoing user/tool messages and keeps only the text. Switching back to a vision-capable model (Claude, GPT-4o/5, Gemini) re-sends the image normally because the session file still stores it.

For a custom-compatible endpoint (mirror, gateway, self-host):

zot --provider deepseek --base-url https://my-deepseek-mirror.example.com/v1 --api-key "$DEEPSEEK_API_KEY"

Google Gemini

zot has built-in Google Gemini support through the AI Studio Generative Language API.

zot --provider google

By default this uses:

model: gemini-2.5-pro
base URL: https://generativelanguage.googleapis.com

Catalog ships with gemini-2.5-pro, gemini-2.5-flash, gemini-2.5-flash-lite, gemini-2.0-flash, and gemini-2.0-flash-lite. Live discovery against /v1beta/models adds anything else your key can see.

Credential lookup order for Google:

--api-key
GEMINI_API_KEY
GOOGLE_API_KEY
$ZOT_HOME/auth.json

Use /login and pick api key to paste an AI Studio key. zot probes /v1beta/models once and stores the key under google in auth.json.

Auth model: API key only (this provider). Google does not issue OAuth tokens for consumer Gemini Advanced / Google One AI Premium subscriptions, so there is no "log in with your Google subscription" flow. The /login subscription step quietly downgrades to the api-key form when you pick Google so you don't end up in a dead end. If you need OAuth/service-account auth instead of an API key, use the google-vertex provider below.

Free-tier rate limits. AI Studio's free tier has tight per-minute and per-day caps that vary by model: gemini-2.5-pro is the strictest (a few requests per minute, ~50 per day), Flash and Flash-Lite are far more generous. If a Pro turn 429s with "You exceeded your current quota" while Flash on the same key still works, you've hit the Pro free-tier RPD. Either switch to Flash for agent loops, or enable billing on your AI Studio project to flip the same key from free to pay-as-you-go pricing ($1.25/M input, $10/M output for Pro).

Reasoning levels (--reasoning off|minimum|low|medium|high|xhigh|max, also configurable through /reasoning or in /settings as reasoning level) map differently per generation. max is a distinct opt-in tier above xhigh. GPT-5.6 and adaptive-thinking Claude models receive native max; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the thinkingLevel enum (MINIMAL/LOW/MEDIUM/HIGH), with Gemini-3-Pro pinned to LOW minimum and HIGH for any medium-or-higher request. off sends no reasoning config. Gemini 2.0 models have no thinking config.

You can add additional Gemini model IDs to models.json under the google provider.

Gemini Enterprise Agent Platform (formerly Google Vertex AI)

zot also has built-in support for Gemini Enterprise Agent Platform, Google's enterprise/GCP-hosted Gemini endpoint. Unlike the AI Studio google provider above, the google-vertex provider supports a Google Cloud API key plus service_account and authorized_user credential files. It does not support consumer Google login or the complete ADC credential chain, such as metadata-server and workload-identity credentials.

zot --provider google-vertex

Required configuration (env vars, read at construction time):

GOOGLE_CLOUD_PROJECT — required, your GCP project id.
GOOGLE_CLOUD_LOCATION — optional, defaults to us-central1.

Credential lookup order for Vertex:

GOOGLE_CLOUD_API_KEY: simplest option. An API key created in the GCP console is sent as x-goog-api-key, without a token exchange.
GOOGLE_APPLICATION_CREDENTIALS: path to a supported credential JSON file.
Default ADC file: if neither of the above is set, zot checks the file written by gcloud auth application-default login. This is ~/.config/gcloud/application_default_credentials.json on Unix systems and %APPDATA%\gcloud\application_default_credentials.json on Windows.

Two credential file shapes are supported:

type: "service_account": zot signs a JWT with the private key and exchanges it for a short-lived access token.
type: "authorized_user": zot exchanges the stored client ID, client secret, and refresh token for a short-lived access token.

Access tokens are cached in memory and refreshed on demand.

If none of these are available, zot errors with vertex: no auth — set GOOGLE_CLOUD_API_KEY or GOOGLE_APPLICATION_CREDENTIALS.

Local models with ollama

zot works with ollama out of the box. Ollama serves an OpenAI-compatible API locally, so any model you have pulled works with zot.

Quick start:

ollama pull qwen3.5:4b
zot --provider ollama --model qwen3.5:4b

That's it. No API key needed for local models. zot defaults to http://localhost:11434.

For a remote ollama instance or one behind auth:

zot --provider ollama --model llama3 --base-url https://my-server.com/v1 --api-key my-token

You can also add models to your models.json so you don't need flags every time:

{
  "providers": {
    "ollama": {
      "models": [
        {
          "id": "qwen3.5:4b",
          "name": "Qwen 3.5 4B",
          "contextWindow": 32768,
          "maxTokens": 8192
        }
      ]
    }
  }
}

The ollama provider uses the OpenAI chat completions protocol internally, so it also works with any OpenAI-compatible server (vLLM, LM Studio, LocalAI, etc.).

Local models with llama.cpp router mode

zot can connect to a recent llama.cpp router, manage its GGUF files, and use loaded models through the router's OpenAI-compatible inference API. This is separate from Ollama. An Ollama server normally listens on port 11434 and should use zot's ollama provider instead.

Install or update llama.cpp. On macOS with Homebrew:

brew install llama.cpp
# or, when already installed
brew upgrade llama.cpp

Create a directory for GGUF files and start llama-server without --model, -m, or -hf. Supplying one of those options starts a single model rather than the router API zot needs.

mkdir -p ~/llama-models

llama-server \
  --models-dir ~/llama-models \
  --no-models-autoload \
  --jinja \
  --host 127.0.0.1 \
  --port 8080 \
  -ngl 999 \
  -c 32768

--no-models-autoload leaves load decisions to /llama. --jinja enables model chat templates and improves tool-call compatibility. -ngl 999 requests maximum GPU offload, while -c 32768 limits each loaded model to a 32K context. Adjust these values for your hardware.

Confirm that router mode is active before configuring zot:

curl http://127.0.0.1:8080/health
curl http://127.0.0.1:8080/models

The models request must return JSON with a data array. A 404 usually means the server is outdated, running on a different port, or was started in single-model mode.

In zot, run /login, choose api key, and select llama.cpp. Enter http://127.0.0.1:8080 as the router URL and leave the API key empty for a local-only server. Do not enter /v1; zot derives the inference URL itself. The saved URL and optional key live in $ZOT_HOME/auth.json.

You can configure the same connection through environment variables:

export LLAMA_BASE_URL=http://127.0.0.1:8080
export LLAMA_API_KEY=optional-secret

When using a key, launch the server with the matching --api-key value. Keep --host 127.0.0.1 unless remote clients must reach the router.

Run /llama to:

inspect the router's current model states
search Hugging Face for GGUF repositories
choose a quantization and download it with byte progress
explicitly load or unload a model
press d to ask the router to remove a downloaded cache model after confirmation

Models discovered through --models-dir or a preset cannot be removed by zot. Delete those files from their configured source instead. The router removes the selected GGUF from its Hugging Face cache, but some llama.cpp versions retain shared repository artifacts such as mmproj files. Remove the repository's cache directory manually if those artifacts are no longer needed. Hugging Face search uses HF_TOKEN when available. Gated repositories require prior access approval, and the llama-server process also needs an authorized HF_TOKEN because the server performs the download.

Opening /model refreshes the router and lists every loaded model under provider llama.cpp. Unloaded models are intentionally omitted because they cannot answer inference requests. Load them through /llama first, then select them through /model.

A model installed through Ollama is kept in Ollama's internal storage and is not automatically available as a llama.cpp GGUF file. Download a GGUF copy through /llama or place GGUF files in ~/llama-models, then restart the router so it discovers files added manually.

Inline images

When a tool returns an image (for example read on a PNG), zot renders it inline on terminals that support it: Ghostty, Kitty, iTerm2, WezTerm. On other terminals you see a text placeholder with MIME type, pixel dimensions, and byte size. Control with the ZOT_INLINE_IMAGES env var:

Value	Effect
unset (default)	Auto-detect based on `TERM_PROGRAM`; use the text placeholder inside VS Code and Herdr.
`iterm`, `iterm2`	Force the iTerm2 OSC 1337 protocol.
`kitty`	Force the Kitty graphics protocol.
`off`, `none`	Always use the text placeholder.

Herdr's Kitty graphics support is currently experimental, so zot does not enable inline images there automatically. After enabling experimental.kitty_graphics in Herdr, set ZOT_INLINE_IMAGES=kitty to opt in.

Frames containing images are full-repainted (no differential diff) to prevent stale image pixels from lingering through scroll. That costs one terminal flash per image-containing frame; set ZOT_INLINE_IMAGES=off if that bothers you.

Tool rendering

By default each tool call (bash, read, write, edit) renders inside a bordered panel — a ┌─ header ─┐, │-prefixed body rows, and a └─┘ footer. On a screen with many calls the borders can read as busy, so zot also offers a flat mode: a single quiet header line per call (▌ bash …) with indented, border-free output. Same information — tool name, arg summary, streamed output, the ... (N more lines, ctrl+o to expand) truncation — just no frame.

Set the tool_render key in $ZOT_HOME/config.json:

{
  "tool_render": "flat"
}

Value	Effect
unset / `"box"` (default)	Each tool call is wrapped in a bordered panel.
`"flat"`	Boxless: a quiet header line plus indented output.

The ZOT_FLAT_TOOLS env var overrides the config for a single run, which is handy for trying it without editing the file:

Value	Effect
`1`, `true`, `yes`, `on`, `flat`	Force flat rendering.
`0`, `false`, `no`, `off`, `box`	Force the bordered panel.
unset	Fall back to the `tool_render` config key.

ZOT_FLAT_TOOLS=1 zot   # flat, just this run
ZOT_FLAT_TOOLS=0 zot   # boxes, even if config.json says "flat"

Either way, theme colors still drive the rendering (the header uses your accent/foreground, output uses the tool-output color) and ctrl+o still expands a truncated result.

Tool arg width

The header line for a tool call shows the tool name plus a one-line summary of its primary argument — a path, a command, or a query. That summary is truncated to 60 cells by default (web_answer What is the best architecture to implement resilience wit...). On a wide terminal that can clip long queries more than you'd like, so set the ZOT_TOOL_ARG_WIDTH env var to raise (or lower) the limit:

ZOT_TOOL_ARG_WIDTH=120 zot   # allow up to 120 cells before truncating

Value	Effect
unset (default)	Truncate the arg summary at 60 cells.
integer in `[20, 500]`	Truncate at that many cells instead.
anything else	Ignored; falls back to the 60-cell default.

Compact input

By default a message you send renders as a padded, background-tinted bubble: a blank tinted row above and below the text, with a ▌ accent bar down the left. So even a one-line prompt occupies three rows. Set compact_input to collapse it to a single quiet ▌ your text gutter line per wrapped row — no padding rows, no background tint.

{
  "compact_input": true
}

Value	Effect
unset / `false` (default)	Padded, background-tinted user bubble.
`true`	One quiet gutter line per wrapped row.

The ZOT_COMPACT_INPUT env var overrides the config for a single run (1/true/on/compact force compact; 0/false/off/bubble force the bubble):

ZOT_COMPACT_INPUT=1 zot   # compact, just this run

Queued messages

You can keep typing while the agent is working. Pressing enter during a turn queues the message instead of interrupting: it shows up above the status bar as sliding in: <text> and is delivered as the next user turn the moment the current one finishes. Queue as many as you want; they run in order. esc cancels the active turn and drops the queue so a runaway turn doesn't flood you with stale follow-ups; ctrl+c while busy arms the exit hint instead of interrupting, a second ctrl+c within two seconds exits zot.

To recover the most recently queued message back into the editor (to tweak it before it runs), press Option+↑. In VS Code's integrated terminal that chord doesn't survive xterm.js's macOS key handling — use Option+Shift+↑ there. zot's hint line under the sliding-in queue adapts automatically based on $TERM_PROGRAM.

Slash commands also work while the agent is busy. Non-destructive ones (/help, /jump, /btw, /sessions, /skills, /reasoning, /settings, /jail, /unjail, /exit) take effect immediately. Destructive ones (/clear, /compact, /login, /logout, /model, /reload-ext) cancel the active turn first and then run.

Keys (interactive mode)

Input

Key	Action
`enter`	Submit (queued if the agent is busy).
`alt+enter`	Newline.
`tab`	Complete the selected slash command.
`esc`	Cancel the current turn (while busy); clear input (while idle).
`ctrl+c`	Clear the input and queue (while idle) or arm the exit hint (while busy). Press again within 2s to exit. Use `esc` to cancel a running turn.
`ctrl+d`	Exit on empty input.
`ctrl+l`	Redraw the screen.
`ctrl+v`	Paste clipboard text into the focused chat, side chat, dialog, filter, or credential input. On Linux this uses `wl-paste`, `xclip`, or `xsel`; terminal-native bracketed paste remains available without those commands. In the main chat on macOS, image-only clipboard content is saved as a temporary PNG and attached to the next prompt.
`ctrl+o`	Expand or collapse long tool results (read, write, edit, bash, glob outputs over ~12 lines).
`ctrl+1` ... `ctrl+9`	Switch to the model bound to that quick-model slot (configured in `/settings` -> model shortcuts). No-op while a turn is running.
`@`	Open the file picker. Browse files and directories in the working directory.

File picker (`@`)

Key	Action
`@`	Open the file picker (type after a space or at the start of input).
`up`, `down`	Navigate the file list.
`right`	Open the selected directory.
`left`	Go back to the parent directory.
`enter`	Select the file or directory and insert it as a chip (`[file:name]` or `[dir:name/]`).
`esc`	Close the file picker.

Type @ followed by a filter string to narrow the list (e.g. @read shows only entries containing "read"). Selected files are inserted as compact chips that expand to the full path on submit. Dragged-and-dropped files and directories also collapse to chips automatically.

Key	Action
`ctrl+a`, `ctrl+e`	Jump to start or end of line.
`alt+left`, `alt+right`	Jump one word back or forward.
`ctrl+u`, `ctrl+k`	Delete to start or end of line.
`ctrl+w`, `alt+backspace`	Delete the previous word.
`up`, `down`	Move within multi-line input. At the top edge, `up` recalls previous prompts and `down` moves forward through prompt history.

Chat scroll

Key	Action
`pgup`, `pgdn`	Scroll one page up or down.
`up`, `down` (editor empty, not browsing prompt history)	Scroll three lines up or down. This is how the mouse wheel reaches the scroll logic on most terminals.

Extensions

zot can be extended in any language via a subprocess + JSON-RPC protocol. Extensions can register slash commands, expose tools to the model, intercept tool calls (block or rewrite args), gate whole turns before the model is called, and rewrite the assistant's visible text before it reaches the user. None are installed by default; opt in explicitly. Hot-reload any time with /reload-ext.

Install and manage

zot ext install <path|git-url>   # copy / clone into $ZOT_HOME/extensions/
zot ext list                      # show installed extensions
zot ext doctor                    # diagnose load, registration, and conflict issues
zot ext logs <name> [-f]          # cat or tail the extension's stderr log
zot ext enable <name>             # re-enable a disabled extension
zot ext disable <name>            # disable without removing
zot ext remove <name>             # delete an extension directory

zot ext doctor keeps normal extension startup fail-soft, but gives you an explicit troubleshooting view: manifest errors, disabled or shadowed extensions, subprocess load errors, ready/auto-ready status, registered commands/tools, registration conflicts, warnings, and the stderr log path.

For development, point zot --ext <path> at a working directory and skip the install step entirely. Repeatable; takes precedence over installed extensions of the same name.

Updating extensions

zot update refreshes the zot binary and every installed extension that lives in a git checkout. Per-extension behaviour:

Disabled extensions are skipped.
Extensions without a .git/ directory (installed by zot ext install ./local-path) are skipped — there is no remote to pull from.
For the rest, zot stashes any dirty worktree state (including untracked runtime files like todos.json or config.json), runs git pull --ff-only, and pops the stash. If the pop produces conflicts, the conflict markers are left in place and you'll see a warning.
Diverged branches, offline pulls, or any other git failure are reported as failed and the next extension is processed. zot update itself never aborts because of an extension.
zot does not run any build step (go build, npm install, make) after the pull. Extension authors are expected to commit the runnable artifact (binary, transpiled JS, etc.). If you need a build, rebuild manually and use /reload-ext.

Theme-only extensions

An extension may ship only a theme: extension.json plus theme.json (or themes/theme.json) and no executable. zot loads it without spawning a subprocess and shows it in /settings with source information. See docs/themes.md.

Reference

examples/extensions/ ships reference implementations in Go, TypeScript, Node, and shell. See docs/extensions.md for the full protocol, the SDK API (packages/agent/ext), and the phase roadmap.

Skills

A skill is a per-folder SKILL.md file with a YAML frontmatter header. zot discovers skills at startup, surfaces their names in the system prompt, and exposes a built-in skill tool the model uses to load the body on demand.

By default zot loads built-in skills plus user-installed skills from:

./.zot/skills/<name>/SKILL.md (project)
$ZOT_HOME/skills/<name>/SKILL.md (global)
./.claude/skills/<name>/SKILL.md, ~/.claude/skills/<name>/SKILL.md (Claude-compatible layout)
./.agents/skills/<name>/SKILL.md, ~/.agents/skills/<name>/SKILL.md (agent-compatible layout)

See docs/skills.md for the frontmatter fields, authoring tips, and example skills under examples/skills/.

Telegram bot (bridge)

zot can run as a telegram bot so you can DM it from your phone. Two ways to run it: from inside the TUI (the running session mirrors into Telegram) or as a standalone background daemon (a headless bot with its own independent agent).

From inside the TUI

Type /telegram in the running TUI to open a picker with connect, disconnect, and status. When connected:

DMs from the paired user become prompts in the same session you're typing in, so you can continue a conversation from the terminal on your phone and back again.
Messages you type in the TUI are mirrored into the Telegram thread prefixed you: ... and the assistant's replies come back prefixed zot: ..., so the Telegram chat stays a complete record of both sides of the conversation.
Messages sent from Telegram show up as your own bubble in Telegram (no mirror) and the assistant's reply to them comes back bare (no prefix).
The status bar shows a - tg - tag while the bridge is active.
/telegram connect / /telegram disconnect / /telegram status (or /tg) also work as direct commands without the picker.

The in-TUI bridge refuses to start while the standalone daemon (below) is running, since two concurrent long-poll consumers of the same bot race on every update and silently drop messages.

Standalone daemon

For headless servers or long-running bots unattached to a TUI:

zot telegram-bot setup     # paste a BotFather token, verify, save
zot telegram-bot run       # foreground: long-poll in this terminal (ctrl+c to stop)
zot telegram-bot start     # background: detach and return immediately
zot telegram-bot stop      # SIGTERM the background bot (SIGKILL after 5s)
zot telegram-bot logs -f   # tail $ZOT_HOME/logs/bot.log (omit -f to just cat)
zot telegram-bot status    # config (token masked) + running/stopped
zot telegram-bot reset     # forget the token and paired user
# short alias: `zot tg ...` is accepted for every subcommand

The background flavor writes the child's PID to $ZOT_HOME/bot.pid and redirects stdout and stderr to $ZOT_HOME/logs/bot.log. zot telegram-bot stop reads that PID, sends SIGTERM, waits up to five seconds, then escalates to SIGKILL if the child is still alive. Running two instances at once is refused at startup.

Use the installed binary for start. go run ./cmd/zot telegram-bot start won't work. go run builds a binary in a temp directory and deletes it when it exits, which kills the detached child. Run make install (or go build) first and invoke the installed binary.

Setup flow:

Talk to @BotFather on telegram, run /newbot, copy the token it gives you.
Run zot telegram-bot setup and paste the token when prompted.
Run zot telegram-bot run in the directory you want the agent to operate in.
Open your bot on telegram, send /start. The first user to do this claims the bridge (stored as allowed_user_id); every other user is rejected.

From then on, any DM you send is forwarded to the agent as a user prompt. Attached photos or image/* documents are downloaded and passed to vision-capable models. In-bot telegram commands are case-insensitive: /help, /status, /stop (cancel the current turn). Config lives in $ZOT_HOME/bot.json (mode 0600).

Starting either Telegram bridge removes any webhook configured for that bot before long polling begins, while preserving pending updates. Telegram does not allow webhooks and getUpdates polling at the same time, so do not share the bot token with another service that expects to keep a webhook active.

Bot mode respects the usual zot flags: --provider, --model, --cwd, --reasoning, --continue, --no-session, --no-tools, and so on. Run zot tg run -c --model claude-opus-4-1 to resume the latest session on Opus, for example.

Architecture: protocol-agnostic bot core

The messenger functionality is split in two layers. A generic, protocol-agnostic core lives in packages/agent/modes/bot: it owns the turn queue, agent prompting, built-in command dispatch (/start, /help, /status, /stop), status formatting, and per-turn credential refresh. Concrete transports implement the small BotAdapter interface (inbound polling, sending replies, a typing indicator, and optional protocol-specific status text); the Telegram support in packages/agent/modes/telegram is one such adapter.

This means additional messaging backends (Discord, Slack, Signal, and similar) can be added by implementing BotAdapter in a new package and wiring up a subcommand. No changes to the runner, agent, or core are required. Channel IDs are opaque strings owned by the adapter, so the shared runner stays free of protocol-specific types.

Development

make build     # build ./bin/zot
make test      # go test -race ./...
make lint      # go vet + gofmt check
make fmt       # gofmt -w .
make release   # cross-compile linux/darwin/windows on amd64 and arm64

Source layout (single Go module, four packages under packages/):

cmd/zot/                              main()
packages/provider/                    LLM client surface, model catalog, streaming clients
packages/provider/auth/               credential store, api-key probe, oauth, login server
packages/core/                        agent loop, sessions, cost tracking, compaction
packages/tui/                         terminal raw-mode, input parser, editor, renderer, markdown, view
packages/agent/                       cli wiring, arg parsing, system prompt, config
packages/agent/extensions/            extension subprocess manager
packages/agent/extproto/              extension wire-format types
packages/agent/modes/                 interactive tui, print, json, dialogs
packages/agent/modes/bot/             protocol-agnostic bot runner (BotAdapter interface)
packages/agent/modes/telegram/        telegram adapter, api client, daemon
packages/agent/tools/                 read, write, edit, bash, glob, sandbox
packages/agent/skills/                skill discovery, frontmatter parser, skill tool
packages/agent/swarm/                 background subagent runtime
packages/agent/sdk/                   public Go SDK for embedding zot in-process (package sdk)
packages/agent/ext/                   public Go SDK for writing extensions (package ext)

Downstream consumers can depend on individual packages: go get github.com/patriceckhart/zot/packages/core pulls only core and its transitive deps (today: provider), no agent or TUI code.

License

MIT

source

1<div align="center">
2  <a href="https://www.zot.sh">
3    <img src="packages/provider/auth/assets/zot-logo.png" alt="zot coding agent harness" width="130" height="130" />
4  </a>
5</div>
6<br>
7<p align="center">
8  <a href="LICENSE"><img alt="license" src="https://img.shields.io/badge/license-MIT-blue"></a>
9  <img alt="Go 1.25+" src="https://img.shields.io/badge/Go-1.25+-00ADD8?logo=go&logoColor=white">
10  <img alt="30+ providers" src="https://img.shields.io/badge/providers-30+-34E2EA">
11</p>
12<p align="center">
13  <a href="https://www.zot.sh">zot.sh</a>
14</p>
15
16## What is it?
17
18Yet another coding agent harness, lightweight and written (vibe-slopped) in go.
19
20- one static binary.
21- built-in providers for Anthropic, OpenAI/Codex/Responses, Kimi, DeepSeek, Google Gemini/Vertex, GitHub Copilot, Bedrock, Azure OpenAI, OpenRouter, Groq, Cerebras, xAI, Together, Hugging Face, Mistral, Moonshot, Z.AI, Xiaomi, MiniMax, Fireworks, Vercel AI Gateway, OpenCode, Cloudflare AI, and Ollama/local models.
22- built-in tools (read, write, edit, bash, glob).
23- three run modes (interactive tui, print, json).
24- built-in telegram bot.
25- extensions in any language via subprocess + json-rpc. None installed by default; opt in with `zot ext install` or `zot --ext`. See [docs/extensions.md](docs/extensions.md).
26- user and extension themes via JSON; see [docs/themes.md](docs/themes.md).
27- standing instructions via `AGENTS.md` files (global and per-project); see [Persistent instructions](#persistent-instructions-agentsmd).
28- reusable instructions via `SKILL.md` files; see [docs/skills.md](docs/skills.md).
29- portable agents from local directories, `.zot` files, or temporary public GitHub downloads; see [docs/zotfiles.md](docs/zotfiles.md).
30
31## Install
32
33### One-liner (macOS, Linux)
34
35```bash
36curl -fsSL https://www.zot.sh/install.sh | bash
37```
38
39Detects your OS and architecture, downloads the latest release from GitHub, verifies the SHA-256 against the release's `checksums.txt`, extracts the binary, and drops it in `/usr/local/bin`, `~/.local/bin`, or `~/bin`, whichever is writable first. Pass a version or prefix to pin:
40
41```bash
42curl -fsSL https://www.zot.sh/install.sh | bash -s -- v0.0.1 ~/bin
43```
44
45### One-liner (Windows, PowerShell)
46
47```powershell
48iwr -useb https://www.zot.sh/install.ps1 | iex
49```
50
51Drops `zot.exe` into `$HOME\bin` and adds it to the user PATH if missing. Open a fresh terminal afterwards.
52
53### go install
54
55```bash
56go install github.com/patriceckhart/zot/cmd/zot@latest
57```
58
59The installed binary reports the tagged module version and supports `zot update`.
60
61### From source
62
63```bash
64git clone https://github.com/patriceckhart/zot
65cd zot
66make build        # produces ./bin/zot
67make install      # into $GOPATH/bin
68```
69
70### Prebuilt binaries
71
72Every release on the [releases page](https://github.com/patriceckhart/zot/releases) ships archives for Linux, macOS, and Windows on amd64 and arm64 (except windows/arm64), plus a `checksums.txt` file. Download, verify, `chmod +x`, and drop on your `$PATH`.
73
74## Authenticate
75
76The easiest way is to just run `zot` and type `/login`. The TUI opens even without credentials and walks you through a browser-based login flow.
77
78### Credential lookup order
79
801. `--api-key` flag
812. provider-specific env var (`ANTHROPIC_API_KEY`, `OPENAI_API_KEY`, `KIMI_API_KEY`, `MOONSHOT_API_KEY`, `DEEPSEEK_API_KEY`, `GEMINI_API_KEY`, `GOOGLE_API_KEY`, `GROQ_API_KEY`, `OPENROUTER_API_KEY`, `MISTRAL_API_KEY`, `XAI_API_KEY`, `CEREBRAS_API_KEY`, `TOGETHER_API_KEY`, `HF_TOKEN`, `ZAI_API_KEY`, `XIAOMI_API_KEY`, `MINIMAX_API_KEY`, `FIREWORKS_API_KEY`, `AI_GATEWAY_API_KEY`, `COPILOT_GITHUB_TOKEN`, `GITHUB_COPILOT_TOKEN`, and others for provider-specific backends)
823. `$ZOT_HOME/auth.json` (API key or OAuth token; mode 0600)
83
84`$ZOT_HOME` defaults to:
85- All platforms: `$XDG_STATE_HOME/zot` when `XDG_STATE_HOME` is set
86- macOS fallback: `~/Library/Application Support/zot`
87- Linux fallback: `~/.local/state/zot`
88- Windows fallback: `%LOCALAPPDATA%\zot`
89
90### API keys from commands
91
92To keep an API key in a password manager instead of `auth.json`, configure an `api_key_command` for the provider:
93
94```json
95{
96  "anthropic": {
97    "api_key_command": {
98      "program": "op",
99      "args": ["read", "op://Work/Anthropic/credential"],
100      "timeout_ms": 120000
101    }
102  }
103}
104```
105
106For a provider added through `models.json`, put the same credential object under `additional_api_key_creds` using its provider ID. `program` is executed directly, without a shell, so each argument must be a separate `args` entry. `timeout_ms` is optional and defaults to 120 seconds.
107
108zot runs the command only when that provider is selected, not while checking login status or refreshing model catalogs in the background. Successful output is cached in memory for the rest of the zot process and is never written to disk. The command must print one non-empty line to stdout; zot removes trailing CR/LF characters, limits output to 64 KiB, and does not include command output in errors. Saving a normal key through `/login` replaces the command configuration, and `/logout` removes it.
109
110Treat `auth.json` as executable configuration: anyone who can modify it can cause zot to run a program under your user account. zot does not interpret `!` prefixes or execute command strings through a shell.
111
112### `/login` flow
113
114Run `zot` and type `/login`. Pick one of two methods:
115
116- **API key**: a small local web server starts on `127.0.0.1:<free-port>`, your browser opens a form, you pick a provider from the full API-key provider list, paste the key, and zot saves it to `auth.json` if accepted. Providers with a lightweight model-list endpoint are probed before saving; provider backends that need extra project/account env vars are saved directly.
117- **Subscription**: use your Claude Pro/Max, ChatGPT Plus/Pro, Kimi Code, SuperGrok/X Premium, or GitHub Copilot subscription. DeepSeek and Google Gemini do **not** have a subscription login path. For those, use the API-key flow.
118  - Anthropic and OpenAI pin the browser callback to fixed provider-specific ports (`localhost:53692` for Anthropic, `localhost:1455` for OpenAI) because those are the only ports their auth servers will redirect to.
119  - Anthropic uses the Claude Code OAuth flow. Messages go to `api.anthropic.com` with a bearer token and the Claude Code identity headers.
120  - OpenAI uses the Codex CLI OAuth flow. Messages go to `chatgpt.com/backend-api/codex/responses` with the `chatgpt-account-id` extracted from the returned id_token.
121  - Kimi uses the Kimi Code device-code OAuth flow. zot opens the verification URL, polls until you approve it in the browser, then sends messages to `api.kimi.com/coding/v1` with the Kimi Code identity headers.
122  - xAI uses a device-code OAuth flow. zot opens a prefilled authorization URL, polls for approval, and uses the resulting token with the xAI API.
123  - GitHub Copilot uses GitHub's device-code login flow. zot stores the GitHub access token and exchanges it for short-lived Copilot inference tokens on demand.
124
125> **Note on subscription login.** The OAuth client IDs used are the ones published in Anthropic's Claude Code CLI, OpenAI's Codex CLI, Kimi Code CLI, xAI's device flow, and GitHub Copilot's device-code flow. Reusing them from a third-party tool may be against their terms of service and may be revoked at any time. Use it at your own risk; the API-key flow is the safe default.
126
127### Token refresh
128
129OAuth access tokens are short-lived (Anthropic ~8h, OpenAI ~30d; Kimi, xAI, and GitHub Copilot also use refresh/exchange flows). zot refreshes or exchanges them automatically:
130
131- At every credential lookup, zot checks the stored `expiry` and, if past it (with a 60s safety margin), hits the provider's `oauth/token` endpoint with the stored `refresh_token`, persists the new `access_token`, `refresh_token`, and `expiry` back to `auth.json`, and hands the fresh token to the client.
132- The telegram bridge additionally refreshes once per turn so a bot that runs for days keeps working without manual intervention.
133- If the refresh itself fails (the `refresh_token` was revoked, or the account was logged out everywhere), the error bubbles up to the caller: the TUI shows it in the status line, the bot replies with it in your DM. Run `/login` to get a fresh token pair.
134
135All data lives under `$ZOT_HOME`:
136
137```
138$ZOT_HOME/
139├── config.json         # last-used provider/model/theme, saved automatically
140├── auth.json           # api keys and oauth tokens (mode 0600)
141├── sessions/           # jsonl transcripts, one dir per cwd
142├── models-cache.json   # live /v1/models discovery cache (6h ttl)
143├── AGENTS.md           # optional: global instructions appended to the prompt
144├── SYSTEM.md           # optional: replaces the default system prompt
145├── skills/             # optional: user SKILL.md files
146├── themes/             # optional: user theme JSON files
147├── extensions/         # installed extensions, one dir per extension
148└── logs/               # app log files
149```
150
151Drop a `SYSTEM.md` in `$ZOT_HOME` to replace the built-in identity and zot-docs guidance for every run. `--system-prompt` still wins per-invocation. Pass an empty value (`--system-prompt ""`) to intentionally omit the built-in identity. Custom prompts still receive appended instructions and generated context, including `AGENTS.md`, skills, auto-swarm guidance when enabled, and the date/cwd footer. Delete the file to revert to the default.
152
153### HTTP proxy
154
155To route zot-managed HTTP and HTTPS requests through one proxy, add `http_proxy` to `$ZOT_HOME/config.json`:
156
157```json
158{
159  "http_proxy": "http://127.0.0.1:7890"
160}
161```
162
163The setting is applied at startup to both HTTP and HTTPS traffic. Existing `HTTP_PROXY`, `HTTPS_PROXY`, `http_proxy`, and `https_proxy` environment variables take precedence for their corresponding protocol. `NO_PROXY` and `no_proxy` continue to control bypasses. Restart zot after changing the config file. If the URL contains proxy credentials, prefer protected environment variables because `config.json` is not a credential store.
164
165## Persistent instructions (AGENTS.md)
166
167Use `AGENTS.md` to give zot standing instructions that layer **on top of** the default system prompt, without replacing it. This is the friendliest way to shape behavior (for example, taming local models that jump straight to code edits) because it adds guidance rather than taking over the base identity the way `SYSTEM.md` does.
168
169zot discovers `AGENTS.md` files automatically at startup and loads them in this order:
170
1711. `$ZOT_HOME/AGENTS.md` (global, machine-wide instructions that apply to every project).
1722. Every `AGENTS.md` from the filesystem root down to the current working directory. More specific (deeper) files may override earlier ones. This includes `~/AGENTS.md` when the working directory is inside your home directory.
173
174All discovered files are appended to the prompt in that order, so a global baseline can be refined per project. To list the active Zotfile agent, loaded context paths, extensions, and user-installed skills above the interactive transcript, enable **show loaded resources at startup** in `/settings`. The agent section appears only for `zot run`, the setting is off by default, built-in skills are omitted, and the displayed lists are not sent to the model or stored in the session transcript.
175
176For a general, non-project-specific instruction set, put your rules in `$ZOT_HOME/AGENTS.md`, for example:
177
178```markdown
179Treat questions and discussions as requests for explanation. Do not edit files or run tools unless explicitly asked to make a change. Ask before modifying code.
180```
181
182`AGENTS.md` vs the other mechanisms:
183
184| Mechanism | Scope | Effect |
185|---|---|---|
186| `$ZOT_HOME/AGENTS.md` | global, every run | appended to the default prompt |
187| `./AGENTS.md` (and parent dirs) | project | appended to the default prompt |
188| `$ZOT_HOME/SYSTEM.md` | global, every run | replaces the built-in identity and zot-docs guidance |
189| `--append-system-prompt <text>` | single run | appends for one invocation (repeatable) |
190| `--system-prompt <text>` | single run | replaces the built-in identity and zot-docs guidance |
191
192> **Note:** zot does not read a `CLAUDE.md` instruction file. The only Claude-compatible thing it picks up is skills under `.claude/skills/`. If you are migrating from Claude Code, move that content into `AGENTS.md` (global or per-project) and zot will use it.
193
194## Changelog on update
195
196The first time you launch a newer zot binary, the TUI shows the GitHub release notes once in a dismissible overlay. Press any key to close. The version is recorded in `config.json`'s `last_changelog_shown` so the same release notes never reappear. Fresh installs don't see a changelog (no upgrade has happened yet). The fetch is best-effort: a network failure or a missing release page silently skips, with another attempt on the next launch.
197
198## Usage
199
200```bash
201zot                              # interactive tui
202zot "fix the failing test"       # tui, pre-filled prompt
203zot -p "list all go files"       # print final text, exit
204echo "list all go files" | zot   # piped stdin implies print mode
205cat README.md | zot -p "summarize this text" # combine stdin with the prompt
206zot -p --stats stats.json "task" # print final text and write generation stats
207zot --json "refactor main.go"    # newline-delimited json events, exit
208zot --continue                   # resume the most recent session for this cwd
209zot --resume                     # pick a session to resume
210zot --list-models                # show supported models
211zot --help
212```
213
214Print-mode stats contain `provider`, `model`, `prompt_tokens`, `reasoning_tokens`, `generated_output_tokens`, and `elapsed_ms`. Counts cover all model turns triggered by the prompt, including tool loops. Prompt tokens include cache reads and writes. `reasoning_tokens` is `null` when the provider does not report a separate count; in that case `generated_output_tokens` is the provider's total output count and may include reasoning. Elapsed time covers the agent run, not startup and credential resolution. The file is written only after a successful run.
215
216## Flags
217
218| Flag | Description |
219|---|---|
220| `--provider <id>` | Pick the provider (for example `anthropic`, `openai`, `openai-codex`, `kimi`, `google`, `github-copilot`, `groq`, `openrouter`, `amazon-bedrock`, `ollama`; see Providers). |
221| `--model <id>` | Pick the model (see `--list-models`). |
222| `--api-key <key>` | Override the API key. |
223| `--base-url <url>` | Override the provider base URL (tests, self-hosted). |
224| `--insecure` | Skip TLS certificate verification for the explicit `--base-url` endpoint or a `baseUrl` defined for a user model in `models.json` (self-signed local/internal inference servers). Built-in providers, auth, and model discovery keep normal TLS verification. |
225| `--system-prompt <text>` | Replace the default system prompt for this run (also overrides `$ZOT_HOME/SYSTEM.md`; pass `""` for no built-in identity). |
226| `--append-system-prompt <text>` | Append text to the system prompt (repeatable). |
227| `--reasoning off\|minimum\|low\|medium\|high\|xhigh\|max` | Set the reasoning level on supported models (default: off). `max` is a separate opt-in tier above `xhigh`. |
228| `--stats <path>` | With `-p`/`--print`, write generation statistics as JSON. |
229| `-c`, `--continue` | Resume the latest session for this cwd. |
230| `-r`, `--resume` | Pick a session to resume. |
231| `--session <path>` | Resume a specific session file. |
232| `--no-session` | Don't read or write session files. |
233| `--cwd <path>` | Use `<path>` as the working directory. |
234| `--no-tools` | Disable all tools. |
235| `--tools <csv>` | Only enable the listed tools. |
236| `--max-steps <n>` | Cap agent loop iterations (default 50). |
237| `-e`, `--ext <path>` | Load an extension from `<path>` for this run (repeatable; wins against installed extensions of the same name). |
238| `--no-ext` | Skip extension discovery for this run. `--ext` still works on top, so `--no-ext --ext ./x` runs only `x`. |
239| `--no-skill` | Disable all skills, including built-ins. No `skill` tool is registered and the system prompt has no skill manifest. |
240| `--no-context-files`, `-nc` | Disable discovery and loading of global and project `AGENTS.md` files. |
241| `--no-yolo` | Confirm every tool call before it runs (interactive TUI only). A dialog shows the tool name and a one-line preview of its args, while edit calls show the proposed diff in the tool panel, with four choices: yes, yes-always-this-tool-this-session, yes-always-this-session, no. Press `/` while the dialog is focused to run a slash command first; closing its input or child dialog returns to the pending confirmation. Ignored with a stderr warning in print / json / rpc modes, where tools still run freely so scripts and automation keep working. |
242
243## Tools
244
245- `read`: read text files, or inline images (PNG, JPEG, GIF, WebP).
246- `write`: create or overwrite files, making parent directories as needed.
247- `edit`: one or more exact-match replacements in an existing file.
248- `bash`: run a command in the session cwd with merged stdout/stderr and a timeout. On Unix, zot uses `/bin/bash -c` when available, then `bash -c` from `PATH`, and falls back to POSIX `/bin/sh -c` when Bash is unavailable. On Windows, it uses `cmd /C`. macOS ships Bash 3.2 by default, so newer Bash features may be unavailable.
249
250When the sandbox is on (see `/jail`), all four tools refuse paths outside the session cwd.
251
252## Modes
253
254- **Interactive** (default): chat TUI with streaming output, spinner, cost meter, slash commands.
255- **Print**: `zot -p "prompt"` runs the agent to completion and writes only the final assistant text to stdout.
256- **Stream**: `zot --stream "prompt"` runs without the TUI and writes assistant text to stdout as it arrives. Tool activity goes to stderr.
257- **Piped input**: when no mode is specified, piped stdin selects print mode. Print, stream, and JSON modes prepend piped stdin to the positional prompt, separated by a newline. For example, `echo "list all go files" | zot` or `cat README.md | zot -p "summarize this text"`.
258- **JSON**: `zot --json "prompt"` emits one JSON object per agent event to stdout, newline-delimited. The schema is documented in [docs/rpc.md](docs/rpc.md).
259- **RPC**: `zot rpc` runs as a long-lived child process; commands in on stdin, events and responses out on stdout, both as NDJSON. Designed for embedding zot in third-party apps written in any language. See [docs/rpc.md](docs/rpc.md) for the wire schema and `examples/rpc/{python,node,shell,go}` for working clients.
260
261## zotfile agents
262
263A zotfile packages an agent's instructions, skills, requirements, and enforced tool permissions as a shareable agent. Run one from a local directory, a packed `.zot` artifact, a short name, or directly from a public GitHub repository:
264
265```bash
266zot run ./my-agent
267zot run ./my-agent.zot
268zot run frkr/zot-archify
269zot run acme/agents/code-reviewer --cwd /path/to/project
270```
271
272Single-part names resolve only to a matching local directory or `.zot` archive. Anyone can publish an agent as a public GitHub repository and run it as `owner/repository`, or publish a collection and run an agent directory as `owner/repository/agent`. zot has no built-in owner, official collection, configured registry, or allowlist. For GitHub sources, zot downloads the repository archive into a temporary directory, validates and runs the selected agent, then removes the downloaded source when the command exits. Agent data, consent receipts, and sessions still persist under `$ZOT_HOME`. See [docs/zotfiles.md](docs/zotfiles.md) for authoring, permissions, packaging, and current limitations.
273
274## Embedding
275
276Two ways to drive zot from another program:
277
278- **Go in-process**: import `github.com/patriceckhart/zot/packages/agent/sdk`. One `Runtime` per project; `Prompt(ctx, text, images)` returns a channel of `Event`. Small example in `examples/sdk/`.
279- **Any language, out-of-process**: spawn `zot rpc` as a subprocess and exchange newline-delimited JSON over its stdin/stdout. Wire format and event schema in [docs/rpc.md](docs/rpc.md). Reference clients live under `examples/rpc/`.
280
281Both interfaces share the same event schema, so transcripts captured by one can be replayed through the other.
282
283## Slash commands
284
285Slash command names are case-insensitive in the TUI and messaging backends; arguments keep their original case. Type `/` in the TUI to open the autocomplete popup. Available commands:
286
287| Command | Description |
288|---|---|
289| `/help` | Show key bindings and commands. |
290| `/login` | Log in via API key or subscription (opens a dialog). |
291| `/logout [provider]` | Clear credentials for any logged-in provider, or all when omitted. `/logout openai-codex` clears ChatGPT/Codex subscription auth while preserving a public OpenAI API key; `/logout kimi` also disables fallback to the official Kimi Code CLI token until you log in to Kimi through zot again. |
292| `/model` | Pick a model from a list (or `/model <id>` to set directly). |
293| `/reasoning` | Set the reasoning level for subsequent model calls. |
294| `/llama` | Connect to the configured llama.cpp router, load, unload, or remove cached models, and search/download GGUF models from Hugging Face with live progress. Shown after llama.cpp login is configured. |
295| `/sessions` | Resume a previous session for this directory. |
296| `/session` | Five ops on the current session: inspect its `timeline`, `export` to a portable `.zotsession` file, `import` one back in, `fork` from a past user message, or view its branch `tree`. Opens a picker without an argument; direct forms include `/session timeline`, `/session export [path]`, `/session import <path>`, `/session fork`, and `/session tree`. Default export destination is `~/Downloads`. |
297| `/jump` | Scroll the chat to a previous turn (or `/jump <text>` to filter). |
298| `/btw` | Side chat with full context that doesn't add to the main thread. |
299| `/swarm` | Spawn, monitor, and chat with background subagents. Each runs in parallel with your main session and shares its working directory. |
300| `/skills` | List discovered skills (SKILL.md files) and preview their bodies. |
301| `/compact` | Summarize the transcript into one message to free up context. |
302| `/study` | Run the canned prompt "Read and understand everything in the current directory." so the agent has full project context before you start asking targeted questions. Pass a path — typed, drag-dropped, or selected via `@` — to target a specific file or directory instead: `/study [dir:packages/]`, `/study cmd/zot/main.go`. |
303| `/jail` | Confine tools to the current directory. |
304| `/unjail` | Allow tools to touch paths outside again. |
305| `/reload-ext` | Hot-reload all extensions (re-read manifests, respawn subprocesses, rebuild tool registry). |
306| `/telegram` | Connect, disconnect, or show status of the Telegram bridge (takes `connect` / `disconnect` / `status` as an optional argument; opens a picker without one). When connected, DMs from the paired user become prompts in the running session and the assistant's replies are mirrored back to Telegram. Alias: `/tg`. |
307| `/settings` | Change persistent settings, including inline images, auto-swarm, and the auto-compact threshold. Saved to `$ZOT_HOME/config.json`; takes effect immediately. |
308| `/clear` | Clear the chat transcript. |
309| `/exit` | Exit zot. |
310
311Extension-registered commands appear under a divider at the bottom of the popup, sorted by name.
312
313### Shell escape (`!command`)
314
315Type `!` followed by a command to run it directly without going through the model. Everything after the `!` is passed to the same shell the `bash` tool uses (`/bin/bash -c` when available on Unix, then `bash -c` from `PATH`, with POSIX `/bin/sh -c` as a fallback; `cmd /C` on Windows), runs in the session working directory, and honors the `/jail` sandbox. The command, merged output, and exit status are appended to the transcript as user context, so the model can use them on the next turn. Running the command does not itself start a model turn. A running `!command` shares the busy state with the agent: `esc` cancels it, and you cannot start one while a turn (or another shell escape) is in flight.
316
317### `/sessions`
318
319Shows previous sessions for the current working directory, newest first, with timestamp, model, message count, cost, and the first user prompt. Pick one with `up`/`down`, `enter` to resume, `esc` to cancel. zot swaps the current session file for the selected one and replays the full transcript (including tool calls) into the agent. Sessions remember the model they ended on, so resuming picks up on that exact model even if your global default changed.
320
321### `/session`
322
323Five ops on the current session. `/session` alone opens a picker; each is also runnable directly.
324
325- **`/session timeline`**. Replaces the chat area with a read-only timeline of the current system prompt, user and assistant messages, and tool calls. The header shows provider-reported context usage and an estimated split between system prompt, tool definitions, and messages. The split uses a simple byte estimate, so it is directional rather than tokenizer-exact. Use `up` / `down` or `pgup` / `pgdn` to select an event, `tab` / `shift+tab` to inspect its summary, payload, result, schema, and transcript-derived timing, and `/` to search. `ctrl+e` writes a JSON export to `~/Downloads` (falling back to the home directory) with `0600` permissions on platforms that support Unix file modes. Images are represented by MIME type and byte size instead of embedding their data. Press `esc` to return to chat.
326- **`/session export [path]`**. Writes the running transcript to a portable `.zotsession` file. Default destination is `~/Downloads/<timestamp>-<session-id>-<prompt-slug>.zotsession`. Pass a path to override; a directory is fine (a dated name is built inside), a bare name gets `.zotsession` appended. The meta's cwd is stripped on the way out so the recipient doesn't see your filesystem layout.
327
328  **What's included.** Only the main chat thread of the running session — messages, tool calls, tool results, compactions, and usage. **`/swarm` subagents are NOT included.** Their transcripts, unix-socket inboxes, and per-agent session files are all machine-local; a `.zotsession` is just a chat transcript and has no way to revive a unix socket on another box. If you want the conversation, copy it out of the dashboard manually.
329- **`/session import <path>`**. Copies a `.zotsession` file into `$ZOT_HOME/sessions/<cwd-hash>/` with a fresh id and the current cwd, then switches the running agent onto it. Imported sessions are first-class: they show up in `/sessions`, `/jump`, and the tree. Drag-drop paths in the editor are accepted (zot strips the surrounding quotes automatically).
330- **`/session fork`**. Opens a turn picker (same shape as `/jump`). Pick any past user message; zot copies every message up to and including that turn into a new session, records `parent` + `fork_point` in the new meta, and switches onto the branch. The parent session stays on disk. Use it to try a different question without polluting the original transcript, or to rewind after the agent went down the wrong path.
331- **`/session tree`**. Shows every session in the current cwd arranged by parent/child relationships, depth-first with indent per level. The current session is tagged `[current]`. Pick any entry to switch into it. Parentless sessions are roots; branches created via `/session fork` nest under whichever session they were forked from. Orphaned children (whose parent file was deleted) still show as roots so they stay discoverable.
332
333### `/jump`
334
335Opens a turn picker for the current session, one row per user prompt, each showing the turn number, how many tools that turn invoked, and the first line of the prompt. `up`/`down` to pick, `enter` to jump, `esc` to cancel. Any printable rune while the picker is open extends a filter; backspace narrows it back. `/jump <text>` pre-applies the filter; if exactly one turn matches, zot jumps straight there without showing the picker.
336
337Jumping is non-destructive. The transcript is untouched, the viewport just scrolls so the chosen turn is at the top. A muted line at the top of the chat reads `viewing turn N of M, pgdn to catch up`. Scroll back to the bottom with `pgdn` (or keep scrolling with the arrow keys) and the indicator goes away.
338
339### `/btw`
340
341Opens a side-chat overlay with the full main session as frozen context, so you can ask quick clarifying questions ("does asyncio.gather() catch exceptions?", "btw the bundle budget is 10MB", "what's the default fetch timeout?") without bloating the main thread.
342
343Each question runs an isolated agent turn against `system + main transcript + side-chat history so far`. The side chat has the same tools and guards as the main chat, including per-tool confirmation when `--no-yolo` is active. Responses, tool calls, and tool results stay in the overlay. When you press `esc` to close, **nothing** has been added to the main session and subsequent main-thread turns don't re-read any of the side-chat exchanges, keeping the running context window lean.
344
345```
346/btw                              # open the overlay, type questions interactively
347/btw does PUT replace the whole resource?
348```
349
350Inside the overlay: `enter` sends, `esc` cancels an in-flight call (or closes the overlay if idle), `ctrl+c` closes immediately. Side-chat exchanges never touch the transcript and aren't persisted to the session file.
351
352### `/swarm`
353
354Background subagents that run alongside your main session. Each one is a separate `zot` subprocess with its own model loop, its own persistent session file, and its own chat in the dashboard — but they all run in **the same working directory as the host**, so they see and edit the same files you do. Spawn one for a side task (“draft the migration”, “investigate this stack trace”, “write the test harness for module X”), keep going in the main thread, check in on it whenever you want.
355
356> **Agents edit the same files you do.** They use the same `read` / `write` / `edit` / `bash` tools as the main agent against the host's working directory. There's no per-agent worktree or branch. If you need parallel edits on isolated checkouts, set that up yourself with `git worktree` outside zot.
357
358```
359/swarm                            # open the dashboard
360/swarm new <task>                 # spawn an agent
361/swarm new --model gpt-5 <task>    # pin the new agent to a specific model
362/swarm logs <id>                  # jump straight into one agent's transcript
363/swarm send <id> <text>           # send a follow-up without opening the dashboard
364/swarm resume                     # pick a stopped agent to bring back
365/swarm resume <id>                # bring a specific agent back
366/swarm kill <id>                  # stop a running agent (its state stays)
367/swarm remove <id>                # delete the agent's session and state
368/swarm list                       # alias for opening the dashboard
369```
370
371**Dashboard (`/swarm` with no arg)** — a list of every agent for the current session, with status, age, and current activity. Keys:
372
373| Key | Action |
374|---|---|
375| `↑` / `↓` | Move cursor between rows. |
376| `enter` | Open the highlighted agent's transcript view. |
377| `n` | Spawn a new agent (opens an inline task editor; inherits the host's current model). |
378| `p` | One-off prompt editor for the selected row (without entering the transcript). |
379| `R` | Resume a stopped agent in place. |
380| `k` | Kill the selected running agent. Its session and state stay so you can resume it later. |
381| `r` | Remove the selected agent entirely (session + meta gone). |
382| `esc` | Close the dashboard. |
383
384**Inside an agent's transcript** — a chat overlay with an always-on inline composer at the bottom. The conversation flows above it; type and `enter` to send a follow-up. The view auto-follows streaming output and shows an inline spinner with the agent's current activity (`thinking`, `tool: edit_file`, etc.) while it's busy. `esc` returns to the dashboard.
385
386**Switching the spawn model from inside the editor** — while composing a task in the `n`-prompt, type `/model` on its own line and `enter`. The standard `/model` picker pops up; pick a model, the picker closes, and the editor reopens with your typed task intact and the new model pinned for the spawn.
387
388**Session scoping** — each agent is stamped with the host session that spawned it and only shows up in that session's dashboard. Swap sessions with `/sessions` and the dashboard re-narrows accordingly. Agents from other sessions keep running in the background and reappear when you switch back.
389
390**Persistence across zot restarts** — every spawn writes a `meta.json` next to its event log and session file under `$ZOT_HOME/swarm/agents/<id>/`. On the next `zot` launch they show up in the dashboard as **detached**; press `R` (or `/swarm resume <id>`) to bring one back. Resumed agents reattach to the same session and inbox socket, so the conversation continues from where it left off.
391
392**Where state lives** — everything per-agent (session file, events log, inbox socket, meta) lives under `$ZOT_HOME/swarm/agents/<id>/`. The agent's actual code edits land directly in your repo; track them with normal `git status` / `git diff`.
393
394**`/session export` does NOT bundle subagents.** A `.zotsession` is just the main chat transcript; per-agent state (session file, unix-socket inbox) is machine-local and doesn't round-trip through a JSONL file. To share what an agent said, copy it out of the transcript view manually.
395
396**Auto-swarm.** With `/settings` -> auto-swarm on, the main agent gets a built-in `swarm_spawn` tool and a system-prompt nudge to use it. It can then fork sub-agents on its own when a request naturally splits into independent parallel work ("implement A and B", "investigate three files"). Each spawn returns the sub-agent id immediately and the main turn keeps going. When every sub-agent the agent spawned in that batch finishes its initial task, zot injects one `[auto-swarm update]` message back into the main chat recapping each agent's status, task, and complete final response. If an agent produced no assistant response, zot includes a truncated transcript tail for diagnostics instead. The main agent then writes a short follow-up summary referencing the agents by id. Off by default; toggle from `/settings`.
397
398### `/settings`
399
400Opens a dialog with every persistent setting. `up`/`down` to navigate, `enter` or `space` to change the selected row, `esc` to close (rows that open a sub-view, like model shortcuts, use `esc` to go back one level first). Changes are written to `$ZOT_HOME/config.json` and take effect on the next turn (no restart needed). Current settings:
401
402- **render images when supported** — draw screenshots / `read`-returned images inline using the terminal's image protocol, or fall back to a text placeholder. Auto-detected from `TERM_PROGRAM`; the toggle overrides the detection. The row is greyed out and forced off on terminals that don't speak any image protocol.
403- **auto-swarm** — let the main agent spawn background sub-agents in parallel via a built-in `swarm_spawn` tool. Off by default. When on, the tool is registered with the running agent, the system prompt gains a short addendum telling the model to delegate independent sub-tasks proactively, and zot watches every sub-agent the main agent spawns. As soon as the last sub-agent in a batch finishes its initial task, an `[auto-swarm update]` message is injected back into the chat with each agent's status / task / transcript tail, so the main agent can summarise the collective outcome. Flipping off mid-session removes the tool from the live agent and strips the addendum on the next turn — the model stops trying to delegate. See `/swarm` for the dashboard that lets you monitor, message, kill, or remove the spawned agents.
404- **auto-compact threshold** — choose `off`, `70%`, `80%`, `85%` (default), or `90%` of the model's advertised context window. The selected percentage controls automatic compaction before and after interactive turns and persists as `auto_compact_threshold`. `off` disables percentage-based triggers but keeps manual `/compact` and automatic recovery from context-window and payload-too-large responses.
405- **jail new sessions by default**: start every new agent with tools confined to its working directory. Off by default. The setting applies to interactive, print, JSON, RPC, and background-agent runs, persists as `jail_by_default`, and immediately updates the current interactive session. `/jail` and `/unjail` remain session-scoped overrides and do not change this default.
406- **OpenRouter Server Tools**: opt in to OpenRouter's beta web search, web fetch, advisor, subagent, fusion, image generation, datetime, and model-search tools. Off by default and available only with the OpenRouter provider. zot limits the remote loop to four server-tool calls per turn. `--no-tools` always disables them. Presets keep their own tool configuration and are not overridden by this setting. Server tools run remotely without zot's local tool confirmation dialog and may incur additional OpenRouter charges.
407- **compact transcript rendering**: reduce visual chrome in the chat transcript. Tool calls render as a quiet header plus indented output instead of a bordered box, and sent messages render without padded background bubbles. Off by default. Changes apply immediately and persist to `config.json` as `compact_mode`.
408- **show loaded resources at startup**: list the active Zotfile agent (for `zot run`), loaded `AGENTS.md` paths, extensions, and user-installed skills in compact sections above the transcript. Built-in skills are omitted. Off by default. Changes apply immediately and persist to `config.json` as `show_instructions_at_startup`.
409- **TUI settings**: opens a sub-view for input layout and status placement. **Input style** can be `plain` (default prompt line), `lines` (separator lines above and below the input), or `block` (a user-bubble-style input block). **Status position** places model, usage, and working-directory information above or below the input. **Working spinner position** places the busy spinner above or below the input. Changes apply immediately and persist to `config.json` as `tui_input_style`, `tui_status_position`, and `tui_working_position` (`above_input` or `below_input` for the position fields).
410- **reasoning level**: choose reasoning for supported models: off, minimum, low, medium, high, xhigh, or max. The `max` tier is opt-in and sent natively to GPT-5.6 and adaptive-thinking Claude models; unsupported backends clamp it to their highest accepted effort. The change is persisted to `config.json` and applied to the next model call. Use `/reasoning` to open this selector directly. The selector only shows distinct levels supported by the active model; models without reasoning support only offer `off`.
411- **color theme** — choose the built-in auto/dark/light theme or any JSON theme discovered under `$ZOT_HOME/themes` or a loaded extension. Theme files can override any subset of UI colors, syntax colors, and spinner frames/messages. Changes apply immediately; if a selected theme file is deleted, zot resets to auto. See [docs/themes.md](docs/themes.md).
412- **model shortcuts** — opens a sub-view with nine slots (`model 1` ... `model 9`). `enter` on a slot opens the same `/model` selector and binds the chosen provider/model to that slot; `backspace` clears a slot. Once assigned, press `Ctrl+1` ... `Ctrl+9` from the editor to switch the active model instantly (the same cross-provider swap `/model` performs, transcript and cost carried over). Assigning a shortcut does not change the current model. Shortcuts are skipped while a turn is running.
413
414### `/skills`
415
416Opens a picker listing every discovered SKILL.md file, built-ins hidden. Each row shows the skill name, source, and description. `enter` opens the body inline (scrollable with `up`/`down`/`pgup`/`pgdn`); `esc` goes back. Re-runs discovery each time it opens, so edits to a SKILL.md during a session are reflected immediately.
417
418### `/compact`
419
420Sends the current transcript through the model with a structured summarization prompt. The returned summary replaces the transcript as one synthetic user message, with the last few exchanges kept verbatim for continuity. The status bar's context meter resets. Use it when the context meter creeps past ~80%.
421
422zot also auto-compacts in the background: after any turn that reaches the configured context threshold, the agent kicks off a condense pass on its own. Choose `off`, `70%`, `80%`, `85%` (default), or `90%` under `/settings` → **auto-compact threshold**. You'll see `condensing history, esc to cancel` above the status bar and an `(auto)` tag next to the context percentage; `esc` aborts it without touching the transcript. Turning the percentage trigger off does not disable automatic compaction and retry after a context-window or payload-too-large response.
423
424### `/jail`
425
426Enforces a sandbox rooted at the cwd shown in the status bar. `read`, `write`, and `edit` resolve their target path (including through symlinks) and refuse anything outside the sandbox. `bash` refuses obvious escape patterns (`sudo`, `rm -rf /`, leading `cd /`, `cd ..`, `cd ~`, `chmod -R`, `dd of=/`, and similar) and rejects shell arguments or redirections that point outside the sandbox. The status bar shows `jailed, ~/your/cwd` while active. Enable **jail new sessions by default** in `/settings` to persist this behavior across launches; `/unjail` then unlocks only the current session.
427
428This is a guardrail against accidents, not a hard security boundary. If you need real isolation, run zot under docker or a proper sandbox.
429
430## Sessions
431
432Every interactive or print/json run (unless `--no-session`) writes a JSONL transcript under `$ZOT_HOME/sessions/<cwd-hash>/`. Resume any of them with `--continue`, `--resume`, `--session <path>`, or interactively via `/sessions` inside the TUI. Empty sessions (the user exited without prompting) are deleted on close so the list stays tidy.
433
434Use `zot sessions prune` outside the TUI to find sessions whose recorded working directories no longer exist. The command groups sessions by directory, shows the number and human-readable total size of stored sessions, lets you select groups, and requires confirmation before permanently deleting files. It preserves sessions when a directory check fails with an error other than "not found" and rechecks each selected directory immediately before deletion. A deleted directory and a path hidden by some unmounted filesystems are indistinguishable, so review the selection and ensure remote filesystems are mounted before deleting missing-directory groups.
435
436Use `--older-than` to prune sessions by last activity instead, including sessions for directories that still exist. Supported age units are `m` (minutes), `h` (hours), `d` (24-hour days), `w` (7-day weeks), `mo` (calendar months), and `y` (calendar years). Add `--cwd PATH` to limit age pruning to one working directory. The session file is checked again immediately before deletion and preserved if it has become active recently.
437
438```sh
439zot sessions prune --dry-run                         # list missing-directory groups
440zot sessions prune --older-than 4h --dry-run         # inspect sessions inactive for four hours
441zot sessions prune --older-than 1mo --cwd ~/project  # select old sessions for one directory
442zot sessions prune --older-than 1y --all              # select every match, then confirm
443zot sessions prune --older-than 30d --all --yes       # delete every match without prompting
444```
445
446`--all` selects every match and still asks for confirmation. Adding `--yes` makes deletion non-interactive and requires `--all`. Malformed, unreadable, symlinked, and non-absolute session entries are reported and preserved during missing-directory pruning. The scan includes normal sessions and named-agent sessions below `$ZOT_HOME/sessions/`; explicit session files stored elsewhere and swarm-agent state are not included.
447
448## Providers
449
450zot's built-in provider catalog includes:
451
452- **Subscription-capable**: Anthropic Claude Pro/Max (`anthropic`), OpenAI Codex / ChatGPT Plus/Pro (`openai-codex`), Kimi Code (`kimi`), SuperGrok/X Premium (`xai`), GitHub Copilot (`github-copilot`).
453- **Direct API providers**: Anthropic, OpenAI Chat Completions, OpenAI Responses, DeepSeek, Google Gemini, Kimi/Moonshot, Moonshot CN, Groq, Cerebras, xAI, Together AI, Hugging Face Router, OpenRouter, Mistral, Z.AI, Xiaomi/MiMo token-plan regions, MiniMax global/CN, Fireworks, Vercel AI Gateway, OpenCode/OpenCode Go.
454- **Cloud/platform providers**: Amazon Bedrock, Google Vertex AI, Azure OpenAI, Cloudflare Workers AI, Cloudflare AI Gateway.
455- **Local/compatible**: Ollama, llama.cpp router mode, and OpenAI-compatible local endpoints via `--base-url`.
456
457Use `/login` to store API keys or subscription credentials. `/model` only shows models from providers that are currently available from env vars, `auth.json`, Kimi CLI fallback, local Ollama, or a configured llama.cpp router.
458
459## Models
460
461`--list-models` or the `/model` picker shows the full catalog across all built-in providers. Three sources:
462
463- **Catalog**: models baked into zot, covering Claude, GPT/Codex, Gemini/Gemma, Kimi/Moonshot, DeepSeek, Groq-hosted Llama/Gemma/Compound, OpenRouter-routed models, Bedrock model ids, Vertex model ids, Azure OpenAI deployments, Copilot models, and other provider-specific catalog entries.
464- **Live**: IDs discovered from `GET /v1/models` using your stored API key (cached for 6h in `$ZOT_HOME/models-cache.json`, refreshed in the background on startup).
465- **Speculative**: IDs that appear in the upstream generator but aren't live on the public API yet. They'll 404 today and start working the moment the provider ships them.
466
467The context meter in the status line uses the model's advertised context window to show how much of it your last turn consumed.
468
469### Model fallback (rescue)
470
471When a turn fails because of a recoverable provider error — expired token (`401`), permission denied (`403`), rate limit (`429`), provider outage (`502`/`503`/`504`), or a transient network failure — zot opens a **rescue** picker over the chat instead of just painting a red banner.
472
473The picker is the same vertical list / fuzzy filter UI as `/model`, but it only shows models from providers you're currently logged in to (env vars, `auth.json`, Kimi CLI fallback, ollama). The failed model is excluded. Press `↑`/`↓` to choose, `enter` to retry the **same prompt** on the new model, `esc` to dismiss.
474
475Before the actual provider request fires, the OpenAI / Anthropic / Kimi / DeepSeek / Google / OpenAI-Codex clients also do up to two silent retries with short backoff (250ms, 750ms) on `502`/`503`/`504` and connection-reset / EOF-before-headers errors. Most edge-proxy blips disappear without you ever seeing the rescue picker.
476
477A rescue retry always **drops launch-time `--api-key` and `--base-url`** before rebuilding the agent. Those overrides are usually the reason the rescue triggered (bad key, typo'd base URL, corporate gateway only valid for the originally-picked provider), so the retry re-resolves credentials from env vars / `auth.json` / provider defaults instead. Use `/model` if you want overrides to stick.
478
479No configuration is required — the candidate list is built dynamically from your active credentials. Bad-request / context-length / serialization errors are NOT routed to the rescue picker, because switching models won't fix them; those still surface as a normal error.
480
481### Custom models
482
483Place a `models.json` in `$ZOT_HOME` (`$XDG_STATE_HOME/zot/` when set, otherwise the platform default above) to add models that aren't in the baked-in catalog or to override existing entries:
484
485```json
486{
487  "providers": {
488    "openai": {
489      "models": [
490        {
491          "id": "gpt-5.5",
492          "name": "GPT-5.5",
493          "reasoning": true,
494          "reasoningLevelMap": {"minimum": "low", "max": ""},
495          "contextWindow": 400000,
496          "maxTokens": 128000
497        }
498      ]
499    }
500  }
501}
502```
503
504Supported fields per model: `id` (required), `name`, `reasoning`, `reasoningLevelMap`, `contextWindow`, `maxTokens`, `baseUrl`, `priceInput`, `priceOutput`, `priceCacheRead`, `priceCacheWrite`.
505
506`reasoningLevelMap` is optional. Protocol defaults apply when it is omitted. Add only model-specific exceptions using `minimum`, `low`, `medium`, `high`, `xhigh`, or `max` as keys. Map a key to another level when both inputs are equivalent, use an identity mapping such as `"max": "max"` to enable a level beyond the protocol default, or map it to an empty string or `off` to remove it. The same effective mapping drives `/reasoning` and provider requests.
507
508Provider keys are normalized: `openai-codex` and `openai-responses` map to `openai`, `anthropic-messages` maps to `anthropic`, `moonshot`, `moonshot-ai`, and `kimi-code` map to `kimi`, and `deepseek-chat` and `deepseek-ai` map to `deepseek`. Built-in provider ids such as `groq`, `openrouter`, `github-copilot`, `amazon-bedrock`, `google-vertex`, `azure-openai-responses`, `fireworks`, `vercel-ai-gateway`, `mistral`, and `xai` can also be used directly.
509
510User-defined models show `source: user` in `--list-models` and take precedence over both the baked-in catalog and live-discovered models. Adding a `models.json` does not hide the built-in catalog; entries are merged on top of it. Missing or invalid files are silently ignored.
511
512#### Custom providers
513
514A top-level provider key that is not a built-in id defines a custom provider. Give it a provider-level `baseUrl` and an `api` wire format (`openai` for OpenAI-compatible Chat Completions, the default, `openai-responses` for the OpenAI Responses API, or `anthropic` for the Anthropic Messages API). A model-level `baseUrl` overrides the provider-level one for that model; an unknown `api` value falls back to `openai` with a warning.
515
516```json
517{
518  "providers": {
519    "my-company": {
520      "baseUrl": "https://llm.mycompany.com/v1",
521      "api": "openai",
522      "models": [
523        { "id": "company-llm-v2", "name": "Company LLM v2" }
524      ]
525    }
526  }
527}
528```
529
530Custom providers are first-class: they appear in `--list-models`, `/model`, and `/login`. `models.json` never stores secrets. Supply the key through `/login`, `--api-key`, or a derived environment variable in upper snake case (so `my-company` reads `MY_COMPANY_API_KEY`). Because many self-hosted gateways do not expose a model-list endpoint, custom provider keys are accepted and stored without a verification probe; an invalid key surfaces on the first model call.
531
532To retrieve this custom provider's key from a password manager, add a matching entry to `$ZOT_HOME/auth.json`:
533
534```json
535{
536  "additional_api_key_creds": {
537    "my-company": {
538      "api_key_command": {
539        "program": "op",
540        "args": ["read", "op://Work/OpenAI/credential"],
541        "timeout_ms": 120000
542      }
543    }
544  }
545}
546```
547
548The provider IDs in `models.json` and `auth.json` must match. Then select the custom model directly:
549
550```bash
551zot --provider my-company --model company-llm-v2
552```
553
554### Kimi Code
555
556zot has built-in Kimi support through the Kimi Coding endpoint and Moonshot's OpenAI-compatible chat API.
557
558```bash
559zot --provider kimi
560```
561
562By default this uses:
563
564- model: `kimi-for-coding`
565- base URL: `https://api.kimi.com/coding/v1`
566
567Credential lookup order for Kimi:
568
5691. `--api-key`
5702. `KIMI_API_KEY`
5713. `MOONSHOT_API_KEY`
5724. `$ZOT_HOME/auth.json`
5735. the official Kimi Code CLI token at `~/.kimi/credentials/kimi-code.json`, unless disabled by `/logout kimi`
574
575Use `/login` for either API-key login or Kimi Code subscription login. The subscription flow uses Kimi Code's device-code OAuth flow: zot opens the verification URL, waits for browser approval, stores the token in `auth.json`, and refreshes it automatically.
576
577For direct Moonshot API keys or a custom compatible endpoint:
578
579```bash
580zot --provider kimi --model kimi-k2-0905-preview --base-url https://api.moonshot.ai/v1 --api-key "$KIMI_API_KEY"
581```
582
583Kimi K3 is built in as `kimi/k3`, `moonshotai/kimi-k3`, `moonshotai-cn/kimi-k3`, `opencode-go/kimi-k3`, `openrouter/moonshotai/kimi-k3`, and `vercel-ai-gateway/moonshotai/kimi-k3`. Its output limit is 131,072 tokens on every built-in route.
584
585You can add additional Kimi/Moonshot model IDs to `models.json` under the `kimi` provider.
586
587### xAI
588
589xAI supports either `XAI_API_KEY` or `/login` subscription authentication. The subscription option is labeled `Sign in with SuperGrok or X Premium`, opens a prefilled device-authorization URL, and refreshes the stored token automatically. The default xAI model is `grok-4.5`, sent through the Responses API.
590
591### DeepSeek
592
593zot has built-in DeepSeek support through DeepSeek's OpenAI-compatible chat API.
594
595```bash
596zot --provider deepseek
597```
598
599By default this uses:
600
601- model: `deepseek-v4-pro`
602- base URL: `https://api.deepseek.com/v1`
603
604Catalog ships with `deepseek-v4-pro` (reasoning) and `deepseek-v4-flash`. These are exactly the IDs returned by `GET https://api.deepseek.com/models` today. You can add additional model IDs to `models.json` under the `deepseek` provider.
605
606Credential lookup order for DeepSeek:
607
6081. `--api-key`
6092. `DEEPSEEK_API_KEY`
6103. `$ZOT_HOME/auth.json`
611
612Use `/login` and pick **api key** to paste a DeepSeek key. zot probes `/v1/models` once and stores the key under `deepseek` in `auth.json`.
613
614> **Auth model: API key only.** DeepSeek does not offer a subscription OAuth flow. The `/login subscription` step lists only Anthropic, OpenAI, and Kimi; DeepSeek shows up only under `/login → api key`.
615
616> **Text only at the wire level.** DeepSeek's chat-completions endpoint currently rejects the multimodal content schema (`unknown variant image_url, expected text`). When the active provider is `deepseek`, zot silently drops `ImageBlock` parts from outgoing user/tool messages and keeps only the text. Switching back to a vision-capable model (Claude, GPT-4o/5, Gemini) re-sends the image normally because the session file still stores it.
617
618For a custom-compatible endpoint (mirror, gateway, self-host):
619
620```bash
621zot --provider deepseek --base-url https://my-deepseek-mirror.example.com/v1 --api-key "$DEEPSEEK_API_KEY"
622```
623
624### Google Gemini
625
626zot has built-in Google Gemini support through the [AI Studio Generative Language API](https://aistudio.google.com/).
627
628```bash
629zot --provider google
630```
631
632By default this uses:
633
634- model: `gemini-2.5-pro`
635- base URL: `https://generativelanguage.googleapis.com`
636
637Catalog ships with `gemini-2.5-pro`, `gemini-2.5-flash`, `gemini-2.5-flash-lite`, `gemini-2.0-flash`, and `gemini-2.0-flash-lite`. Live discovery against `/v1beta/models` adds anything else your key can see.
638
639Credential lookup order for Google:
640
6411. `--api-key`
6422. `GEMINI_API_KEY`
6433. `GOOGLE_API_KEY`
6444. `$ZOT_HOME/auth.json`
645
646Use `/login` and pick **api key** to paste an AI Studio key. zot probes `/v1beta/models` once and stores the key under `google` in `auth.json`.
647
648> **Auth model: API key only (this provider).** Google does not issue OAuth tokens for consumer Gemini Advanced / Google One AI Premium subscriptions, so there is no "log in with your Google subscription" flow. The `/login subscription` step quietly downgrades to the api-key form when you pick Google so you don't end up in a dead end. If you need OAuth/service-account auth instead of an API key, use the `google-vertex` provider below.
649
650> **Free-tier rate limits.** AI Studio's free tier has tight per-minute and per-day caps that vary by model: `gemini-2.5-pro` is the strictest (a few requests per minute, ~50 per day), Flash and Flash-Lite are far more generous. If a Pro turn 429s with `"You exceeded your current quota"` while Flash on the same key still works, you've hit the Pro free-tier RPD. Either switch to Flash for agent loops, or [enable billing](https://aistudio.google.com/app/apikey) on your AI Studio project to flip the same key from free to pay-as-you-go pricing (`$1.25/M` input, `$10/M` output for Pro).
651
652Reasoning levels (`--reasoning off|minimum|low|medium|high|xhigh|max`, also configurable through `/reasoning` or in `/settings` as **reasoning level**) map differently per generation. `max` is a distinct opt-in tier above `xhigh`. GPT-5.6 and adaptive-thinking Claude models receive native `max`; unsupported providers clamp it to their highest accepted effort. Budget-based providers retain their provider/model caps. Gemini 3.x uses the `thinkingLevel` enum (`MINIMAL`/`LOW`/`MEDIUM`/`HIGH`), with Gemini-3-Pro pinned to `LOW` minimum and `HIGH` for any medium-or-higher request. `off` sends no reasoning config. Gemini 2.0 models have no thinking config.
653
654You can add additional Gemini model IDs to `models.json` under the `google` provider.
655
656### Gemini Enterprise Agent Platform (formerly Google Vertex AI)
657
658zot also has built-in support for [Gemini Enterprise Agent Platform](https://cloud.google.com/products/gemini-enterprise-agent-platform), Google's enterprise/GCP-hosted Gemini endpoint. Unlike the AI Studio `google` provider above, the `google-vertex` provider supports a Google Cloud API key plus `service_account` and `authorized_user` credential files. It does not support consumer Google login or the complete ADC credential chain, such as metadata-server and workload-identity credentials.
659
660```bash
661zot --provider google-vertex
662```
663
664Required configuration (env vars, read at construction time):
665
666- `GOOGLE_CLOUD_PROJECT` — required, your GCP project id.
667- `GOOGLE_CLOUD_LOCATION` — optional, defaults to `us-central1`.
668
669Credential lookup order for Vertex:
670
6711. `GOOGLE_CLOUD_API_KEY`: simplest option. An API key created in the GCP console is sent as `x-goog-api-key`, without a token exchange.
6722. `GOOGLE_APPLICATION_CREDENTIALS`: path to a supported credential JSON file.
6733. **Default ADC file**: if neither of the above is set, zot checks the file written by `gcloud auth application-default login`. This is `~/.config/gcloud/application_default_credentials.json` on Unix systems and `%APPDATA%\gcloud\application_default_credentials.json` on Windows.
674
675Two credential file shapes are supported:
676
677- `type: "service_account"`: zot signs a JWT with the private key and exchanges it for a short-lived access token.
678- `type: "authorized_user"`: zot exchanges the stored client ID, client secret, and refresh token for a short-lived access token.
679
680Access tokens are cached in memory and refreshed on demand.
681
682If none of these are available, zot errors with `vertex: no auth — set GOOGLE_CLOUD_API_KEY or GOOGLE_APPLICATION_CREDENTIALS`.
683
684### Local models with ollama
685
686zot works with [ollama](https://ollama.com) out of the box. Ollama serves an OpenAI-compatible API locally, so any model you have pulled works with zot.
687
688Quick start:
689
690```bash
691ollama pull qwen3.5:4b
692zot --provider ollama --model qwen3.5:4b
693```
694
695That's it. No API key needed for local models. zot defaults to `http://localhost:11434`.
696
697For a remote ollama instance or one behind auth:
698
699```bash
700zot --provider ollama --model llama3 --base-url https://my-server.com/v1 --api-key my-token
701```
702
703You can also add models to your `models.json` so you don't need flags every time:
704
705```json
706{
707  "providers": {
708    "ollama": {
709      "models": [
710        {
711          "id": "qwen3.5:4b",
712          "name": "Qwen 3.5 4B",
713          "contextWindow": 32768,
714          "maxTokens": 8192
715        }
716      ]
717    }
718  }
719}
720```
721
722The `ollama` provider uses the OpenAI chat completions protocol internally, so it also works with any OpenAI-compatible server (vLLM, LM Studio, LocalAI, etc.).
723
724### Local models with llama.cpp router mode
725
726zot can connect to a recent [llama.cpp](https://github.com/ggml-org/llama.cpp) router, manage its GGUF files, and use loaded models through the router's OpenAI-compatible inference API. This is separate from Ollama. An Ollama server normally listens on port `11434` and should use zot's `ollama` provider instead.
727
728Install or update llama.cpp. On macOS with Homebrew:
729
730```bash
731brew install llama.cpp
732# or, when already installed
733brew upgrade llama.cpp
734```
735
736Create a directory for GGUF files and start `llama-server` without `--model`, `-m`, or `-hf`. Supplying one of those options starts a single model rather than the router API zot needs.
737
738```bash
739mkdir -p ~/llama-models
740
741llama-server \
742  --models-dir ~/llama-models \
743  --no-models-autoload \
744  --jinja \
745  --host 127.0.0.1 \
746  --port 8080 \
747  -ngl 999 \
748  -c 32768
749```
750
751`--no-models-autoload` leaves load decisions to `/llama`. `--jinja` enables model chat templates and improves tool-call compatibility. `-ngl 999` requests maximum GPU offload, while `-c 32768` limits each loaded model to a 32K context. Adjust these values for your hardware.
752
753Confirm that router mode is active before configuring zot:
754
755```bash
756curl http://127.0.0.1:8080/health
757curl http://127.0.0.1:8080/models
758```
759
760The models request must return JSON with a `data` array. A 404 usually means the server is outdated, running on a different port, or was started in single-model mode.
761
762In zot, run `/login`, choose **api key**, and select **llama.cpp**. Enter `http://127.0.0.1:8080` as the router URL and leave the API key empty for a local-only server. Do not enter `/v1`; zot derives the inference URL itself. The saved URL and optional key live in `$ZOT_HOME/auth.json`.
763
764You can configure the same connection through environment variables:
765
766```bash
767export LLAMA_BASE_URL=http://127.0.0.1:8080
768export LLAMA_API_KEY=optional-secret
769```
770
771When using a key, launch the server with the matching `--api-key` value. Keep `--host 127.0.0.1` unless remote clients must reach the router.
772
773Run `/llama` to:
774
775- inspect the router's current model states
776- search Hugging Face for GGUF repositories
777- choose a quantization and download it with byte progress
778- explicitly load or unload a model
779- press `d` to ask the router to remove a downloaded cache model after confirmation
780
781Models discovered through `--models-dir` or a preset cannot be removed by zot. Delete those files from their configured source instead. The router removes the selected GGUF from its Hugging Face cache, but some llama.cpp versions retain shared repository artifacts such as `mmproj` files. Remove the repository's cache directory manually if those artifacts are no longer needed. Hugging Face search uses `HF_TOKEN` when available. Gated repositories require prior access approval, and the `llama-server` process also needs an authorized `HF_TOKEN` because the server performs the download.
782
783Opening `/model` refreshes the router and lists every loaded model under provider `llama.cpp`. Unloaded models are intentionally omitted because they cannot answer inference requests. Load them through `/llama` first, then select them through `/model`.
784
785A model installed through Ollama is kept in Ollama's internal storage and is not automatically available as a llama.cpp GGUF file. Download a GGUF copy through `/llama` or place GGUF files in `~/llama-models`, then restart the router so it discovers files added manually.
786
787## Inline images
788
789When a tool returns an image (for example `read` on a PNG), zot renders it inline on terminals that support it: **Ghostty**, **Kitty**, **iTerm2**, **WezTerm**. On other terminals you see a text placeholder with MIME type, pixel dimensions, and byte size. Control with the `ZOT_INLINE_IMAGES` env var:
790
791| Value | Effect |
792|---|---|
793| unset (default) | Auto-detect based on `TERM_PROGRAM`; use the text placeholder inside VS Code and Herdr. |
794| `iterm`, `iterm2` | Force the iTerm2 OSC 1337 protocol. |
795| `kitty` | Force the Kitty graphics protocol. |
796| `off`, `none` | Always use the text placeholder. |
797
798Herdr's Kitty graphics support is currently experimental, so zot does not enable inline images there automatically. After enabling `experimental.kitty_graphics` in Herdr, set `ZOT_INLINE_IMAGES=kitty` to opt in.
799
800Frames containing images are full-repainted (no differential diff) to prevent stale image pixels from lingering through scroll. That costs one terminal flash per image-containing frame; set `ZOT_INLINE_IMAGES=off` if that bothers you.
801
802## Tool rendering
803
804By default each tool call (bash, read, write, edit) renders inside a bordered panel — a `┌─ header ─┐`, `│`-prefixed body rows, and a `└─┘` footer. On a screen with many calls the borders can read as busy, so zot also offers a **flat** mode: a single quiet header line per call (`▌ bash …`) with indented, border-free output. Same information — tool name, arg summary, streamed output, the `... (N more lines, ctrl+o to expand)` truncation — just no frame.
805
806Set the `tool_render` key in `$ZOT_HOME/config.json`:
807
808```json
809{
810  "tool_render": "flat"
811}
812```
813
814| Value | Effect |
815|---|---|
816| unset / `"box"` (default) | Each tool call is wrapped in a bordered panel. |
817| `"flat"` | Boxless: a quiet header line plus indented output. |
818
819The `ZOT_FLAT_TOOLS` env var overrides the config for a single run, which is handy for trying it without editing the file:
820
821| Value | Effect |
822|---|---|
823| `1`, `true`, `yes`, `on`, `flat` | Force flat rendering. |
824| `0`, `false`, `no`, `off`, `box` | Force the bordered panel. |
825| unset | Fall back to the `tool_render` config key. |
826
827```sh
828ZOT_FLAT_TOOLS=1 zot   # flat, just this run
829ZOT_FLAT_TOOLS=0 zot   # boxes, even if config.json says "flat"
830```
831
832Either way, theme colors still drive the rendering (the header uses your accent/foreground, output uses the tool-output color) and `ctrl+o` still expands a truncated result.
833
834### Tool arg width
835
836The header line for a tool call shows the tool name plus a one-line summary of its primary argument — a `path`, a `command`, or a query. That summary is truncated to 60 cells by default (`web_answer What is the best architecture to implement resilience wit...`). On a wide terminal that can clip long queries more than you'd like, so set the `ZOT_TOOL_ARG_WIDTH` env var to raise (or lower) the limit:
837
838```sh
839ZOT_TOOL_ARG_WIDTH=120 zot   # allow up to 120 cells before truncating
840```
841
842| Value | Effect |
843|---|---|
844| unset (default) | Truncate the arg summary at 60 cells. |
845| integer in `[20, 500]` | Truncate at that many cells instead. |
846| anything else | Ignored; falls back to the 60-cell default. |
847
848## Compact input
849
850By default a message you send renders as a padded, background-tinted bubble: a blank tinted row above and below the text, with a `▌` accent bar down the left. So even a one-line prompt occupies three rows. Set `compact_input` to collapse it to a single quiet `▌ your text` gutter line per wrapped row — no padding rows, no background tint.
851
852```json
853{
854  "compact_input": true
855}
856```
857
858| Value | Effect |
859|---|---|
860| unset / `false` (default) | Padded, background-tinted user bubble. |
861| `true` | One quiet gutter line per wrapped row. |
862
863The `ZOT_COMPACT_INPUT` env var overrides the config for a single run (`1`/`true`/`on`/`compact` force compact; `0`/`false`/`off`/`bubble` force the bubble):
864
865```sh
866ZOT_COMPACT_INPUT=1 zot   # compact, just this run
867```
868
869## Queued messages
870
871You can keep typing while the agent is working. Pressing `enter` during a turn queues the message instead of interrupting: it shows up above the status bar as `sliding in: <text>` and is delivered as the next user turn the moment the current one finishes. Queue as many as you want; they run in order. `esc` cancels the active turn and drops the queue so a runaway turn doesn't flood you with stale follow-ups; `ctrl+c` while busy arms the exit hint instead of interrupting, a second `ctrl+c` within two seconds exits zot.
872
873To recover the most recently queued message back into the editor (to tweak it before it runs), press `Option+↑`. In VS Code's integrated terminal that chord doesn't survive xterm.js's macOS key handling — use `Option+Shift+↑` there. zot's hint line under the sliding-in queue adapts automatically based on `$TERM_PROGRAM`.
874
875Slash commands also work while the agent is busy. Non-destructive ones (`/help`, `/jump`, `/btw`, `/sessions`, `/skills`, `/reasoning`, `/settings`, `/jail`, `/unjail`, `/exit`) take effect immediately. Destructive ones (`/clear`, `/compact`, `/login`, `/logout`, `/model`, `/reload-ext`) cancel the active turn first and then run.
876
877
878## Keys (interactive mode)
879
880### Input
881
882| Key | Action |
883|---|---|
884| `enter` | Submit (queued if the agent is busy). |
885| `alt+enter` | Newline. |
886| `tab` | Complete the selected slash command. |
887| `esc` | Cancel the current turn (while busy); clear input (while idle). |
888| `ctrl+c` | Clear the input and queue (while idle) or arm the exit hint (while busy). Press again within 2s to exit. Use `esc` to cancel a running turn. |
889| `ctrl+d` | Exit on empty input. |
890| `ctrl+l` | Redraw the screen. |
891| `ctrl+v` | Paste clipboard text into the focused chat, side chat, dialog, filter, or credential input. On Linux this uses `wl-paste`, `xclip`, or `xsel`; terminal-native bracketed paste remains available without those commands. In the main chat on macOS, image-only clipboard content is saved as a temporary PNG and attached to the next prompt. |
892| `ctrl+o` | Expand or collapse long tool results (read, write, edit, bash, glob outputs over ~12 lines). |
893| `ctrl+1` ... `ctrl+9` | Switch to the model bound to that quick-model slot (configured in `/settings` -> model shortcuts). No-op while a turn is running. |
894| `@` | Open the file picker. Browse files and directories in the working directory. |
895
896### File picker (`@`)
897
898| Key | Action |
899|---|---|
900| `@` | Open the file picker (type after a space or at the start of input). |
901| `up`, `down` | Navigate the file list. |
902| `right` | Open the selected directory. |
903| `left` | Go back to the parent directory. |
904| `enter` | Select the file or directory and insert it as a chip (`[file:name]` or `[dir:name/]`). |
905| `esc` | Close the file picker. |
906
907Type `@` followed by a filter string to narrow the list (e.g. `@read` shows only entries containing "read"). Selected files are inserted as compact chips that expand to the full path on submit. Dragged-and-dropped files and directories also collapse to chips automatically.
908
909### Editor line navigation
910
911| Key | Action |
912|---|---|
913| `ctrl+a`, `ctrl+e` | Jump to start or end of line. |
914| `alt+left`, `alt+right` | Jump one word back or forward. |
915| `ctrl+u`, `ctrl+k` | Delete to start or end of line. |
916| `ctrl+w`, `alt+backspace` | Delete the previous word. |
917| `up`, `down` | Move within multi-line input. At the top edge, `up` recalls previous prompts and `down` moves forward through prompt history. |
918
919### Chat scroll
920
921| Key | Action |
922|---|---|
923| `pgup`, `pgdn` | Scroll one page up or down. |
924| `up`, `down` (editor empty, not browsing prompt history) | Scroll three lines up or down. This is how the mouse wheel reaches the scroll logic on most terminals. |
925
926## Extensions
927
928zot can be extended in any language via a subprocess + JSON-RPC protocol. Extensions can register slash commands, expose tools to the model, intercept tool calls (block or rewrite args), gate whole turns before the model is called, and rewrite the assistant's visible text before it reaches the user. None are installed by default; opt in explicitly. Hot-reload any time with `/reload-ext`.
929
930### Install and manage
931
932```bash
933zot ext install <path|git-url>   # copy / clone into $ZOT_HOME/extensions/
934zot ext list                      # show installed extensions
935zot ext doctor                    # diagnose load, registration, and conflict issues
936zot ext logs <name> [-f]          # cat or tail the extension's stderr log
937zot ext enable <name>             # re-enable a disabled extension
938zot ext disable <name>            # disable without removing
939zot ext remove <name>             # delete an extension directory
940```
941
942`zot ext doctor` keeps normal extension startup fail-soft, but gives
943you an explicit troubleshooting view: manifest errors, disabled or
944shadowed extensions, subprocess load errors, ready/auto-ready status,
945registered commands/tools, registration conflicts, warnings, and the
946stderr log path.
947
948For development, point `zot --ext <path>` at a working directory and skip the install step entirely. Repeatable; takes precedence over installed extensions of the same name.
949
950### Updating extensions
951
952`zot update` refreshes the zot binary **and** every installed extension that lives in a git checkout. Per-extension behaviour:
953
954- Disabled extensions are skipped.
955- Extensions without a `.git/` directory (installed by `zot ext install ./local-path`) are skipped — there is no remote to pull from.
956- For the rest, zot stashes any dirty worktree state (including untracked runtime files like `todos.json` or `config.json`), runs `git pull --ff-only`, and pops the stash. If the pop produces conflicts, the conflict markers are left in place and you'll see a warning.
957- Diverged branches, offline pulls, or any other git failure are reported as `failed` and the next extension is processed. `zot update` itself never aborts because of an extension.
958- zot does **not** run any build step (`go build`, `npm install`, `make`) after the pull. Extension authors are expected to commit the runnable artifact (binary, transpiled JS, etc.). If you need a build, rebuild manually and use `/reload-ext`.
959
960### Theme-only extensions
961
962An extension may ship only a theme: `extension.json` plus `theme.json` (or `themes/theme.json`) and no executable. zot loads it without spawning a subprocess and shows it in `/settings` with source information. See [docs/themes.md](docs/themes.md).
963
964### Reference
965
966`examples/extensions/` ships reference implementations in Go, TypeScript, Node, and shell. See [docs/extensions.md](docs/extensions.md) for the full protocol, the SDK API (`packages/agent/ext`), and the phase roadmap.
967
968## Skills
969
970A skill is a per-folder `SKILL.md` file with a YAML frontmatter header. zot discovers skills at startup, surfaces their names in the system prompt, and exposes a built-in `skill` tool the model uses to load the body on demand.
971
972By default zot loads built-in skills plus user-installed skills from:
973
974- `./.zot/skills/<name>/SKILL.md` (project)
975- `$ZOT_HOME/skills/<name>/SKILL.md` (global)
976- `./.claude/skills/<name>/SKILL.md`, `~/.claude/skills/<name>/SKILL.md` (Claude-compatible layout)
977- `./.agents/skills/<name>/SKILL.md`, `~/.agents/skills/<name>/SKILL.md` (agent-compatible layout)
978
979See [docs/skills.md](docs/skills.md) for the frontmatter fields, authoring tips, and example skills under `examples/skills/`.
980
981## Telegram bot (bridge)
982
983zot can run as a telegram bot so you can DM it from your phone. Two ways to run it: **from inside the TUI** (the running session mirrors into Telegram) or **as a standalone background daemon** (a headless bot with its own independent agent).
984
985### From inside the TUI
986
987Type `/telegram` in the running TUI to open a picker with **connect**, **disconnect**, and **status**. When connected:
988
989- DMs from the paired user become prompts in the **same** session you're typing in, so you can continue a conversation from the terminal on your phone and back again.
990- Messages you type in the TUI are mirrored into the Telegram thread prefixed `you: ...` and the assistant's replies come back prefixed `zot: ...`, so the Telegram chat stays a complete record of both sides of the conversation.
991- Messages sent from Telegram show up as your own bubble in Telegram (no mirror) and the assistant's reply to them comes back bare (no prefix).
992- The status bar shows a `- tg -` tag while the bridge is active.
993- `/telegram connect` / `/telegram disconnect` / `/telegram status` (or `/tg`) also work as direct commands without the picker.
994
995The in-TUI bridge refuses to start while the standalone daemon (below) is running, since two concurrent long-poll consumers of the same bot race on every update and silently drop messages.
996
997### Standalone daemon
998
999For headless servers or long-running bots unattached to a TUI:
1000
1001```bash
1002zot telegram-bot setup     # paste a BotFather token, verify, save
1003zot telegram-bot run       # foreground: long-poll in this terminal (ctrl+c to stop)
1004zot telegram-bot start     # background: detach and return immediately
1005zot telegram-bot stop      # SIGTERM the background bot (SIGKILL after 5s)
1006zot telegram-bot logs -f   # tail $ZOT_HOME/logs/bot.log (omit -f to just cat)
1007zot telegram-bot status    # config (token masked) + running/stopped
1008zot telegram-bot reset     # forget the token and paired user
1009# short alias: `zot tg ...` is accepted for every subcommand
1010```
1011
1012The background flavor writes the child's PID to `$ZOT_HOME/bot.pid` and redirects stdout and stderr to `$ZOT_HOME/logs/bot.log`. `zot telegram-bot stop` reads that PID, sends SIGTERM, waits up to five seconds, then escalates to SIGKILL if the child is still alive. Running two instances at once is refused at startup.
1013
1014> **Use the installed binary for `start`.** `go run ./cmd/zot telegram-bot start` won't work. `go run` builds a binary in a temp directory and deletes it when it exits, which kills the detached child. Run `make install` (or `go build`) first and invoke the installed binary.
1015
1016Setup flow:
1017
10181. Talk to [@BotFather](https://t.me/BotFather) on telegram, run `/newbot`, copy the token it gives you.
10192. Run `zot telegram-bot setup` and paste the token when prompted.
10203. Run `zot telegram-bot run` in the directory you want the agent to operate in.
10214. Open your bot on telegram, send `/start`. The first user to do this claims the bridge (stored as `allowed_user_id`); every other user is rejected.
1022
1023From then on, any DM you send is forwarded to the agent as a user prompt. Attached photos or `image/*` documents are downloaded and passed to vision-capable models. In-bot telegram commands are case-insensitive: `/help`, `/status`, `/stop` (cancel the current turn). Config lives in `$ZOT_HOME/bot.json` (mode 0600).
1024
1025Starting either Telegram bridge removes any webhook configured for that bot before long polling begins, while preserving pending updates. Telegram does not allow webhooks and `getUpdates` polling at the same time, so do not share the bot token with another service that expects to keep a webhook active.
1026
1027Bot mode respects the usual zot flags: `--provider`, `--model`, `--cwd`, `--reasoning`, `--continue`, `--no-session`, `--no-tools`, and so on. Run `zot tg run -c --model claude-opus-4-1` to resume the latest session on Opus, for example.
1028
1029### Architecture: protocol-agnostic bot core
1030
1031The messenger functionality is split in two layers. A generic, protocol-agnostic core lives in `packages/agent/modes/bot`: it owns the turn queue, agent prompting, built-in command dispatch (`/start`, `/help`, `/status`, `/stop`), status formatting, and per-turn credential refresh. Concrete transports implement the small `BotAdapter` interface (inbound polling, sending replies, a typing indicator, and optional protocol-specific status text); the Telegram support in `packages/agent/modes/telegram` is one such adapter.
1032
1033This means additional messaging backends (Discord, Slack, Signal, and similar) can be added by implementing `BotAdapter` in a new package and wiring up a subcommand. No changes to the runner, agent, or core are required. Channel IDs are opaque strings owned by the adapter, so the shared runner stays free of protocol-specific types.
1034
1035## Development
1036
1037```bash
1038make build     # build ./bin/zot
1039make test      # go test -race ./...
1040make lint      # go vet + gofmt check
1041make fmt       # gofmt -w .
1042make release   # cross-compile linux/darwin/windows on amd64 and arm64
1043```
1044
1045Source layout (single Go module, four packages under `packages/`):
1046
1047```
1048cmd/zot/                              main()
1049packages/provider/                    LLM client surface, model catalog, streaming clients
1050packages/provider/auth/               credential store, api-key probe, oauth, login server
1051packages/core/                        agent loop, sessions, cost tracking, compaction
1052packages/tui/                         terminal raw-mode, input parser, editor, renderer, markdown, view
1053packages/agent/                       cli wiring, arg parsing, system prompt, config
1054packages/agent/extensions/            extension subprocess manager
1055packages/agent/extproto/              extension wire-format types
1056packages/agent/modes/                 interactive tui, print, json, dialogs
1057packages/agent/modes/bot/             protocol-agnostic bot runner (BotAdapter interface)
1058packages/agent/modes/telegram/        telegram adapter, api client, daemon
1059packages/agent/tools/                 read, write, edit, bash, glob, sandbox
1060packages/agent/skills/                skill discovery, frontmatter parser, skill tool
1061packages/agent/swarm/                 background subagent runtime
1062packages/agent/sdk/                   public Go SDK for embedding zot in-process (package sdk)
1063packages/agent/ext/                   public Go SDK for writing extensions (package ext)
1064```
1065
1066Downstream consumers can depend on individual packages:
1067`go get github.com/patriceckhart/zot/packages/core` pulls only `core` and its transitive deps (today: `provider`), no agent or TUI code.
1068
1069## License
1070
1071MIT

README.md RawHistoryFolder

What is it?

Install

One-liner (macOS, Linux)

One-liner (Windows, PowerShell)

go install

From source

Prebuilt binaries

Authenticate

Credential lookup order

API keys from commands

/login flow

Token refresh

HTTP proxy

Persistent instructions (AGENTS.md)

Changelog on update

Usage

Flags

Tools

Modes

zotfile agents

Embedding

Slash commands

Shell escape (!command)

/sessions

/session

/jump

/btw

/swarm

/settings

/skills

/compact

/jail

Sessions

Providers

Models

Model fallback (rescue)

Custom models

Custom providers

Kimi Code

xAI

DeepSeek

Google Gemini

Gemini Enterprise Agent Platform (formerly Google Vertex AI)

Local models with ollama

Local models with llama.cpp router mode

Inline images

Tool rendering

Tool arg width

Compact input

Queued messages

Keys (interactive mode)

Input

File picker (@)

Editor line navigation

Chat scroll

Extensions

Install and manage

Updating extensions

Theme-only extensions

Reference

Skills

Telegram bot (bridge)

From inside the TUI

Standalone daemon

Architecture: protocol-agnostic bot core

Development

License

README.md Raw History Folder

`/login` flow

Shell escape (`!command`)

`/sessions`

`/session`

`/jump`

`/btw`

`/swarm`

`/settings`

`/skills`

`/compact`

`/jail`

File picker (`@`)