1package modes
2
3import (
4 "context"
5 "encoding/json"
6 "fmt"
7 "io"
8
9 "github.com/patriceckhart/zot/packages/core"
10 "github.com/patriceckhart/zot/packages/provider"
11)
12
13// RunJSON runs the agent to completion, writing one JSON object per
14// AgentEvent as newline-delimited JSON.
15func RunJSON(ctx context.Context, ag *core.Agent, prompt string, images []provider.ImageBlock, out io.Writer) error {
16 enc := json.NewEncoder(out)
17 write := func(v any) {
18 _ = enc.Encode(v)
19 }
20
21 var runErr error
22 sink := func(ev core.AgentEvent) {
23 write(EventToJSON(ev))
24 }
25
26 if err := ag.Prompt(ctx, prompt, images, sink); err != nil {
27 runErr = err
28 }
29
30 if runErr != nil {
31 fmt.Fprintln(out, `{"type":"error","message":`+jsonString(runErr.Error())+`}`)
32 }
33 return runErr
34}
35
36// EventToJSON converts an AgentEvent to a JSON-friendly map. The on-wire
37// schema is deliberately simple and flat. Exported so the RPC mode can
38// reuse the same serialisation as `zot --json`.
39func EventToJSON(ev core.AgentEvent) map[string]any {
40 m := map[string]any{"type": ev.Type()}
41 switch e := ev.(type) {
42 case core.EvTurnStart:
43 m["step"] = e.Step
44 case core.EvUserMessage:
45 m["content"] = ContentToJSON(e.Message.Content)
46 m["time"] = e.Message.Time
47 case core.EvAssistantMessage:
48 m["content"] = ContentToJSON(e.Message.Content)
49 m["time"] = e.Message.Time
50 case core.EvTextDelta:
51 m["delta"] = e.Delta
52 case core.EvToolUseStart:
53 m["id"] = e.ID
54 m["name"] = e.Name
55 case core.EvToolUseArgs:
56 m["id"] = e.ID
57 m["delta"] = e.Delta
58 case core.EvToolUseEnd:
59 m["id"] = e.ID
60 case core.EvToolCall:
61 m["id"] = e.ID
62 m["name"] = e.Name
63 var args any
64 _ = json.Unmarshal(e.Args, &args)
65 m["args"] = args
66 case core.EvToolProgress:
67 m["id"] = e.ID
68 m["text"] = e.Text
69 case core.EvToolResult:
70 m["id"] = e.ID
71 m["is_error"] = e.Result.IsError
72 m["content"] = ContentToJSON(e.Result.Content)
73 case core.EvUsage:
74 m["input"] = e.Usage.InputTokens
75 m["output"] = e.Usage.OutputTokens
76 m["reasoning"] = nullableReasoningTokens(e.Usage)
77 m["cache_read"] = e.Usage.CacheReadTokens
78 m["cache_write"] = e.Usage.CacheWriteTokens
79 m["cost_usd"] = e.Usage.CostUSD
80 m["cumulative"] = map[string]any{
81 "input": e.Cumulative.InputTokens,
82 "output": e.Cumulative.OutputTokens,
83 "reasoning": nullableReasoningTokens(e.Cumulative),
84 "cache_read": e.Cumulative.CacheReadTokens,
85 "cache_write": e.Cumulative.CacheWriteTokens,
86 "cost_usd": e.Cumulative.CostUSD,
87 }
88 case core.EvTurnEnd:
89 m["stop"] = string(e.Stop)
90 if e.Err != nil {
91 m["error"] = e.Err.Error()
92 }
93 }
94 return m
95}
96
97// ContentToJSON serialises a transcript content slice into the same
98// shape used in EventToJSON. Exported alongside EventToJSON for the
99// RPC mode.
100func ContentToJSON(blocks []provider.Content) []map[string]any {
101 out := make([]map[string]any, 0, len(blocks))
102 for _, b := range blocks {
103 switch v := b.(type) {
104 case provider.TextBlock:
105 out = append(out, map[string]any{"type": "text", "text": v.Text})
106 case provider.ImageBlock:
107 out = append(out, map[string]any{"type": "image", "mime_type": v.MimeType, "bytes": len(v.Data)})
108 case provider.ToolCallBlock:
109 var args any
110 _ = json.Unmarshal(v.Arguments, &args)
111 out = append(out, map[string]any{"type": "tool_call", "id": v.ID, "name": v.Name, "args": args})
112 case provider.ToolResultBlock:
113 out = append(out, map[string]any{
114 "type": "tool_result",
115 "call_id": v.CallID,
116 "is_error": v.IsError,
117 "content": ContentToJSON(v.Content),
118 })
119 }
120 }
121 return out
122}
123
124func nullableReasoningTokens(usage provider.Usage) any {
125 if !usage.ReasoningTokensKnown {
126 return nil
127 }
128 return usage.ReasoningTokens
129}
130
131func jsonString(s string) string {
132 b, _ := json.Marshal(s)
133 return string(b)
134}