You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
searchSubagent-<uuid>.jsonl — (optional) search subagent work
33
37
title-<uuid>.jsonl — (optional, UI-only) title generation
@@ -37,6 +41,8 @@ debug-logs/<sessionId>/
37
41
38
42
Always read `main.jsonl` first — it has the full conversation flow. Child files only appear when those operations occurred. `main.jsonl` contains `child_session_ref` entries that link to each child file by name. Title, categorization, and summarize files are UI housekeeping and rarely relevant to troubleshooting. When investigating model availability or selection issues, read `models.json` — it contains the full list of models (with capabilities, billing, and limits) that were available when the session started.
39
43
44
+
When investigating what the model was told (system prompt, instructions), read the `system_prompt_*.json` file referenced by a `system_prompt_ref` entry in `main.jsonl`. The file contains the full untruncated system prompt as `{ "content": "..." }`. When investigating which tools were available, read the `tools_*.json` file similarly. If the model changed mid-session, multiple numbered files exist — each `llm_request` entry has a `systemPromptFile` attr indicating which file was active for that request.
45
+
40
46
Each line is a JSON object. Common fields: `ts` (epoch ms), `dur` (duration ms), `sid` (session ID), `type`, `name`, `spanId`, `parentSpanId`, `status` (`ok`|`error`), `attrs` (type-specific details).
41
47
42
48
### Event Type Reference with Examples
@@ -58,15 +64,15 @@ Key attrs: `args` (JSON string of tool input), `result` (tool output or error te
Key attrs: `model`, `inputTokens`, `outputTokens`, `ttft` (time to first token in ms), `error` (when failed).
69
+
Key attrs: `model`, `inputTokens`, `outputTokens`, `ttft` (time to first token in ms), `maxTokens`, `temperature`, `topP`, `systemPromptFile` (references a file in the session directory), `userRequest` (the full user message content, untruncated), `inputMessages` (full messages array as JSON, pre-truncated at 64KB), `error` (when failed).
64
70
65
71
#### agent_response — model output (text + tool calls)
66
72
```jsonl
67
-
{"ts":1773200234011,"dur":0,"sid":"62f52dec","type":"agent_response","name":"agent_response","spanId":"agent-msg-000000000000000c","parentSpanId":"0000000000000003","status":"ok","attrs":{"response":"[{\"role\":\"assistant\",\"parts\":[{\"type\":\"text\",\"content\":\"Running your command now.\"},{\"type\":\"tool_call\",\"name\":\"run_in_terminal\",\"arguments\":\"{...}\"}]}]"}}
73
+
{"ts":1773200234011,"dur":0,"sid":"62f52dec","type":"agent_response","name":"agent_response","spanId":"agent-msg-000000000000000c","parentSpanId":"0000000000000003","status":"ok","attrs":{"response":"[{\"role\":\"assistant\",...}]","reasoning":"The user wants me to run a command."}}
68
74
```
69
-
Key attrs: `response` (JSON-encoded array of message parts; may be truncated).
75
+
Key attrs: `response` (JSON-encoded array of message parts; may be truncated), `reasoning` (optional — the model's chain-of-thought/thinking text when thinking mode is active; may be truncated).
-`system_prompt_ref` — references a `system_prompt_*.json` file in the session directory. `attrs.file` is the filename, `attrs.model` is the model it was written for. Read this file to see the full system prompt.
96
+
-`tools_ref` — references a `tools_*.json` file. `attrs.file` is the filename, `attrs.model` is the model.
97
+
98
+
#### session_start — session metadata (appears once at session start)
Key attrs: `turnId` (iteration number within a single user request's tool-calling loop). Use these to identify which iteration events belong to and to count total loop iterations.
110
+
88
111
### Reading the event hierarchy
89
112
90
113
Events form a tree via `spanId`/`parentSpanId`. A typical chain:
0 commit comments