[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-posthog-exploring-llm-traces":3,"mdc-163bev-key":44,"related-repo-posthog-exploring-llm-traces":3806,"related-org-posthog-exploring-llm-traces":3924},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":29,"repoUrl":30,"updatedAt":31,"license":32,"forks":33,"topics":34,"repo":39,"sourceUrl":42,"mdContent":43},"exploring-llm-traces","debug and inspect LLM agent traces","ABSOLUTE MUST to debug and inspect LLM\u002FAI agent traces using PostHog's MCP tools. Use when the user pastes a trace or session URL (e.g. \u002Fai-observability\u002Ftraces\u002F\u003Cid> or \u002Fai-observability\u002Fsessions\u002F\u003Cid>), asks to debug a trace, figure out what went wrong, check if an agent used a tool correctly, verify context\u002Ffiles were surfaced, inspect subagent behavior, investigate LLM decisions, or analyze token usage and costs. Also use when raw SQL\u002FHogQL against `events.properties.$ai_input` \u002F `$ai_output_choices` returns empty — message content lives only on the dedicated `posthog.ai_events` table.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"posthog","PostHog","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fposthog.png",[12,16,19,20,23,26],{"name":13,"slug":14,"type":15},"Tracing","tracing","tag",{"name":17,"slug":18,"type":15},"Observability","observability",{"name":9,"slug":8,"type":15},{"name":21,"slug":22,"type":15},"LLM","llm",{"name":24,"slug":25,"type":15},"Agents","agents",{"name":27,"slug":28,"type":15},"Debugging","debugging",59,"https:\u002F\u002Fgithub.com\u002FPostHog\u002Fai-plugin","2026-04-06T18:44:40.791466",null,11,[35,36,37,38],"claude-code-plugin","codex-plugin","cursor-plugin","gemini-cli-extension",{"repoUrl":30,"stars":29,"forks":33,"topics":40,"description":41},[35,36,37,38],"Official PostHog plugin for Claude Code, Cursor, Gemini, Codex and other AI coding tools","https:\u002F\u002Fgithub.com\u002FPostHog\u002Fai-plugin\u002Ftree\u002FHEAD\u002Fskills\u002Fexploring-llm-traces","---\nname: exploring-llm-traces\ndescription: >\n  ABSOLUTE MUST to debug and inspect LLM\u002FAI agent traces using PostHog's MCP tools.\n  Use when the user pastes a trace or session URL (e.g. \u002Fai-observability\u002Ftraces\u002F\u003Cid> or \u002Fai-observability\u002Fsessions\u002F\u003Cid>),\n  asks to debug a trace, figure out what went wrong, check if an agent used a tool correctly,\n  verify context\u002Ffiles were surfaced, inspect subagent behavior, investigate LLM decisions,\n  or analyze token usage and costs. Also use when raw SQL\u002FHogQL against\n  `events.properties.$ai_input` \u002F `$ai_output_choices` returns empty — message content lives only\n  on the dedicated `posthog.ai_events` table.\n---\n\n# Exploring LLM traces with MCP tools\n\nPostHog captures LLM\u002FAI agent activity as traces. Each trace is a tree of events representing\na single AI interaction — from the top-level agent invocation down to individual LLM API calls.\n\n## Available tools\n\n| Tool                            | Purpose                                                       |\n| ------------------------------- | ------------------------------------------------------------- |\n| `posthog:query-llm-traces-list` | Search and list traces; can return large multi-trace payloads |\n| `posthog:query-llm-trace`       | Get a single trace by ID with full event tree                 |\n| `posthog:read-data-schema`      | Discover custom event\u002Fperson properties before filtering      |\n| `posthog:execute-sql`           | Ad-hoc SQL for complex trace analysis                         |\n\n## Event hierarchy\n\nSee the [event reference](.\u002Freferences\u002Fevents-and-properties.md) for the full schema.\n\n```text\n$ai_trace (top-level container)\n  └── $ai_span (logical groupings, e.g. \"RAG retrieval\", \"tool execution\")\n        ├── $ai_generation (individual LLM API call)\n        └── $ai_embedding (embedding creation)\n```\n\nEvents are linked via `$ai_parent_id` → parent's `$ai_span_id` or `$ai_trace_id`.\n\n## Workflow: debug a trace or session from a URL\n\n### Step 1 — Classify the URL\n\nFirst inspect the path. Do not treat every UUID-looking value as a trace ID.\n\n- `\u002Fai-observability\u002Ftraces\u002F\u003Ctrace_id>` or legacy `\u002Fllm-analytics\u002Ftraces\u002F\u003Ctrace_id>` \u002F `\u002Fllm-observability\u002Ftraces\u002F\u003Ctrace_id>` is a single trace. Fetch it with `posthog:query-llm-trace`.\n- `\u002Fai-observability\u002Fsessions\u002F\u003Csession_id>` or legacy `\u002Fllm-analytics\u002Fsessions\u002F\u003Csession_id>` is an AI session, not a trace. Fetch traces with `posthog:query-llm-traces-list` filtered by event property `$ai_session_id`.\n\nPreserve `date_from` \u002F `date_to` query parameters from the URL when present.\nIf none are present but the URL has a `timestamp` query parameter, use that timestamp as the anchor and query an absolute window around it, for example `timestamp - 36h` to `timestamp + 36h`.\nThis handles exact session links whose UI timestamp may be offset from the stored event timestamps while keeping the query bounded.\nIf the URL has neither explicit dates nor `timestamp`, use a safe default like `{\"date_from\": \"-7d\"}`.\n\nFor exact trace and session URLs, skip schema discovery for the standard `$ai_*` fields used below. These are AI observability built-ins, not project-specific custom properties.\n\n### Step 2 — Fetch trace data\n\nFor a trace URL, call `posthog:query-llm-trace` with:\n\n```json\n{\n  \"traceId\": \"\u003Ctrace_id>\",\n  \"dateRange\": { \"date_from\": \"-7d\" }\n}\n```\n\nFor a session URL, call `posthog:query-llm-traces-list` with:\n\n```json\n{\n  \"dateRange\": { \"date_from\": \"\u003Ctimestamp_minus_36h>\", \"date_to\": \"\u003Ctimestamp_plus_36h>\" },\n  \"filterTestAccounts\": false,\n  \"limit\": 20,\n  \"properties\": [{ \"type\": \"event\", \"key\": \"$ai_session_id\", \"value\": [\"\u003Csession_id>\"], \"operator\": \"exact\" }]\n}\n```\n\nUse the URL's `date_from` \u002F `date_to` values in the session query if present.\nIf the URL only has `timestamp`, calculate the absolute date range from that timestamp instead of using a relative range like `-1h`.\nSet `filterTestAccounts: false` for an exact URL so the requested trace is not hidden by account filters.\n\nThe result contains the event tree with all properties.\nThe response may be large — when it exceeds the inline limit, Claude Code auto-persists it to a file.\n\nFrom the result you get:\n\n- Every event with its type (`$ai_span`, `$ai_generation`, etc.)\n- Span names (`$ai_span_name`) — these are the tool\u002Fstep names\n- Latency, error flags, models used\n- Parent-child relationships via `$ai_parent_id`\n- `_posthogUrl` — **always include this in your response** so the user can click through to the UI\n\n### Step 3 — Parse large results with scripts\n\nWhen the result is persisted to a file (large traces with full `$ai_input`\u002F`$ai_output_choices`),\nuse the [parsing scripts](.\u002Fscripts\u002F) to explore it.\n\n**Start with the summary** to get the full picture, then drill into specifics:\n\n```bash\n# 1. Overview: metadata, tool calls, final output, errors\npython3 scripts\u002Fprint_summary.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 2. Timeline: chronological event list with truncated I\u002FO\npython3 scripts\u002Fprint_timeline.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 3. Drill into a specific span's full input\u002Foutput\nSPAN=\"tool_name\" python3 scripts\u002Fextract_span.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 4. Full conversation with thinking blocks and tool calls\npython3 scripts\u002Fextract_conversation.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 5. Search for a keyword across all properties\nSEARCH=\"keyword\" python3 scripts\u002Fsearch_traces.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n```\n\nAll scripts support `MAX_LEN=N` env var to control truncation (0 = unlimited).\n\n## Investigation patterns\n\n### \"Did the agent use the tool correctly?\"\n\n1. Find the `$ai_span` for the tool call (look at `$ai_span_name`)\n2. Check `$ai_input_state` — what arguments were passed to the tool?\n3. Check `$ai_output_state` — what did the tool return?\n4. Check `$ai_is_error` — did the tool call fail?\n\n### \"Was the context correct?\" \u002F \"Were the right files surfaced?\"\n\n1. Find the `$ai_generation` event where the LLM made the decision\n2. Check `$ai_input` — this is the full message history the LLM saw\n3. Look at preceding `$ai_span` events for retrieval\u002Fsearch steps\n4. Check their `$ai_output_state` — what content was retrieved and fed to the LLM?\n\n### \"Did the subagent work?\"\n\n1. In the structural overview, find spans that are children of other spans (via `$ai_parent_id`)\n2. The parent span is the orchestrator; child spans are subagent steps\n3. Check each child's `$ai_output_state` and `$ai_is_error`\n4. If a child span contains `$ai_generation` events, those are the subagent's LLM calls\n\n### \"Why did the LLM say X?\"\n\n1. Use `search_traces.py` to find where the text appears: `SEARCH=\"the text\" python3 scripts\u002Fsearch_traces.py FILE`\n2. This shows which event and property path contains it\n3. Check the `$ai_input` of that generation to see what the LLM was told before it said X\n\n## Constructing UI links\n\nThe trace tools return `_posthogUrl` — always surface this to the user.\n\nYou can also construct links manually:\n\n- **Trace detail**: `https:\u002F\u002Fapp.posthog.com\u002Fai-observability\u002Ftraces\u002F\u003Ctrace_id>?timestamp=\u003Curl_encoded_timestamp>&event=\u003Coptional_event_id>`\n- **Traces list with filters**: returned in `_posthogUrl` from `query-llm-traces-list`\n\nThe `timestamp` query param is **required** — use the `createdAt` of the earliest event in the trace, URL-encoded (e.g. `timestamp=2026-04-01T19%3A39%3A20Z`).\n\nWhen presenting findings, always include the relevant PostHog URL so the user can verify.\n\n## Finding traces\n\nUse `posthog:query-llm-traces-list` to search and filter traces.\n\n**CRITICAL: Never assume event names, property names, or property values from training data.**\nEvery project instruments different custom properties. For open-ended searches and custom filters, call\n`posthog:read-data-schema` first to discover what properties and values actually exist in the project's\ndata before constructing filters.\n\nThe exception is exact AI observability trace\u002Fsession URLs: use the built-in `$ai_trace_id` \u002F `$ai_session_id`\nfields directly and skip schema discovery.\n\n### Discovering the schema first\n\nBefore filtering traces, discover what's available:\n\n1. **Confirm AI events exist** — call `posthog:read-data-schema` with `kind: \"events\"` and look for `$ai_*` events\n2. **Find filterable properties** — call `posthog:read-data-schema` with `kind: \"event_properties\"` and `event_name: \"$ai_generation\"` (or another AI event) to see what properties are captured\n3. **Get actual values** — call `posthog:read-data-schema` with `kind: \"event_property_values\"`, `event_name: \"$ai_generation\"`, and `property_name: \"$ai_model\"` to see real model names in use\n\nOnly then construct the `query-llm-traces-list` call with property filters.\n\nThis is especially important for custom properties like `project_id`, `conversation_id`, `user_tier`, etc. — these vary per project and cannot be guessed.\n\nDo not confirm `$ai_*` properties, but confirm any other like `email` of a person.\n\n### By filters\n\n```json\nposthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"limit\": 20,\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"$ai_model\", \"value\": \"gpt-4o\", \"operator\": \"exact\"}\n  ]\n}\n```\n\nMultiple filters are AND-ed together:\n\n```json\nposthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"$ai_provider\", \"value\": \"anthropic\", \"operator\": \"exact\"},\n    {\"type\": \"event\", \"key\": \"$ai_is_error\", \"value\": [\"true\"], \"operator\": \"exact\"}\n  ]\n}\n```\n\nYou can also filter by person properties (discover them via `read-data-schema` with `kind: \"entity_properties\"` and `entity: \"person\"`):\n\n```json\nposthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"properties\": [\n    {\"type\": \"person\", \"key\": \"email\", \"value\": \"@company.com\", \"operator\": \"icontains\"}\n  ]\n}\n```\n\n### By external identifiers\n\nCustomers often store their own IDs as event or person properties.\nUse `posthog:read-data-schema` to discover what custom properties exist, then filter:\n\n1. Call `posthog:read-data-schema` with `kind: \"event_properties\"` and `event_name: \"$ai_trace\"` to find custom properties\n2. Review the returned properties and their sample values\n3. Construct the filter using the discovered property key and a known value\n\n```json\nposthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-7d\"},\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"project_id\", \"value\": \"proj_abc123\", \"operator\": \"exact\"}\n  ]\n}\n```\n\nFor more complex SQL patterns, read these references:\n\n- [Single trace retrieval](.\u002Freferences\u002Fexample-llm-trace.md.j2) — fetches a single trace by ID with all events and properties (renders the `TraceQuery` HogQL)\n- [Traces list with aggregated metrics](.\u002Freferences\u002Fexample-llm-traces-list.md) — two-phase query: find trace IDs first, then fetch aggregated latency, tokens, costs, and error counts\n\n## Parsing large trace results\n\nTrace tool results are JSON. When too large to read inline, Claude Code persists them to a file.\n\n### Persisted file format\n\n```json\n[{ \"type\": \"text\", \"text\": \"{\\\"results\\\": [...], \\\"_posthogUrl\\\": \\\"...\\\"}\" }]\n```\n\n### Trace JSON structure\n\n```text\nresults (array for list, object for single trace)\n  ├── id, traceName, createdAt, totalLatency, totalCost\n  ├── inputState, outputState (trace-level state)\n  └── events[]\n        ├── event ($ai_span | $ai_generation | $ai_embedding | $ai_metric | $ai_feedback)\n        ├── id, createdAt\n        └── properties\n              ├── $ai_span_name, $ai_latency, $ai_is_error\n              ├── $ai_input_state, $ai_output_state (span tool I\u002FO)\n              ├── $ai_input, $ai_output_choices (generation messages)\n              ├── $ai_model, $ai_provider\n              └── $ai_input_tokens, $ai_output_tokens, $ai_total_cost_usd\n```\n\n### Available scripts\n\n| Script                                                         | Purpose                                                                                 | Usage                                                    |\n| -------------------------------------------------------------- | --------------------------------------------------------------------------------------- | -------------------------------------------------------- |\n| [`print_summary.py`](.\u002Fscripts\u002Fprint_summary.py)               | Aggregate list\u002Fsession totals, trace metadata, tool calls, errors, and final LLM output | `python3 scripts\u002Fprint_summary.py FILE`                  |\n| [`print_timeline.py`](.\u002Fscripts\u002Fprint_timeline.py)             | Chronological event timeline with I\u002FO summaries                                         | `python3 scripts\u002Fprint_timeline.py FILE`                 |\n| [`extract_span.py`](.\u002Fscripts\u002Fextract_span.py)                 | Full input\u002Foutput of a specific span by name                                            | `SPAN=\"name\" python3 scripts\u002Fextract_span.py FILE`       |\n| [`extract_conversation.py`](.\u002Fscripts\u002Fextract_conversation.py) | LLM messages with thinking blocks and tool calls                                        | `python3 scripts\u002Fextract_conversation.py FILE`           |\n| [`search_traces.py`](.\u002Fscripts\u002Fsearch_traces.py)               | Find a keyword across all event properties                                              | `SEARCH=\"keyword\" python3 scripts\u002Fsearch_traces.py FILE` |\n| [`show_structure.py`](.\u002Fscripts\u002Fshow_structure.py)             | Show JSON keys and types without values                                                 | `cat blob.json \\| python3 scripts\u002Fshow_structure.py`     |\n\n## Tips\n\n- Always set `dateRange` — queries without a time range are slow. Use narrow windows (`-30m`, `-1h`) for broad listing queries; wider windows (`-7d`, `-30d`) are fine for narrow queries filtered by trace ID or specific property values\n- Always include the `_posthogUrl` in your response so the user can click through\n- `$ai_input_state` \u002F `$ai_output_state` on spans contain tool call inputs and outputs\n- `$ai_input` \u002F `$ai_output_choices` on generations contain the full LLM conversation — can be megabytes; when the result is persisted to a file, use the parsing scripts\n- In raw SQL, heavy content (`$ai_input` \u002F `$ai_output` \u002F `$ai_output_choices` \u002F `$ai_input_state` \u002F `$ai_output_state` \u002F `$ai_tools`) lives only on the `posthog.ai_events` table, not `events.properties` — see the [event reference](.\u002Freferences\u002Fevents-and-properties.md) for the column mapping and trace-id-anchored query patterns\n- Use `filterTestAccounts: true` to exclude internal\u002Ftest traffic when searching\n- `$ai_trace` events are NOT in the `events` array — their data is surfaced via trace-level `inputState`, `outputState`, and `traceName`\n",{"data":45,"body":46},{"name":4,"description":6},{"type":47,"children":48},"root",[49,58,64,71,168,174,188,200,229,235,242,247,317,375,388,394,406,542,553,889,930,935,940,1011,1017,1046,1056,1272,1285,1291,1297,1357,1363,1412,1418,1466,1472,1511,1517,1529,1534,1576,1611,1616,1622,1633,1650,1668,1674,1679,1781,1793,1820,1840,1846,2152,2157,2570,2597,2871,2877,2889,2927,3176,3181,3214,3220,3225,3231,3364,3370,3379,3385,3590,3596,3800],{"type":50,"tag":51,"props":52,"children":54},"element","h1",{"id":53},"exploring-llm-traces-with-mcp-tools",[55],{"type":56,"value":57},"text","Exploring LLM traces with MCP tools",{"type":50,"tag":59,"props":60,"children":61},"p",{},[62],{"type":56,"value":63},"PostHog captures LLM\u002FAI agent activity as traces. Each trace is a tree of events representing\na single AI interaction — from the top-level agent invocation down to individual LLM API calls.",{"type":50,"tag":65,"props":66,"children":68},"h2",{"id":67},"available-tools",[69],{"type":56,"value":70},"Available tools",{"type":50,"tag":72,"props":73,"children":74},"table",{},[75,94],{"type":50,"tag":76,"props":77,"children":78},"thead",{},[79],{"type":50,"tag":80,"props":81,"children":82},"tr",{},[83,89],{"type":50,"tag":84,"props":85,"children":86},"th",{},[87],{"type":56,"value":88},"Tool",{"type":50,"tag":84,"props":90,"children":91},{},[92],{"type":56,"value":93},"Purpose",{"type":50,"tag":95,"props":96,"children":97},"tbody",{},[98,117,134,151],{"type":50,"tag":80,"props":99,"children":100},{},[101,112],{"type":50,"tag":102,"props":103,"children":104},"td",{},[105],{"type":50,"tag":106,"props":107,"children":109},"code",{"className":108},[],[110],{"type":56,"value":111},"posthog:query-llm-traces-list",{"type":50,"tag":102,"props":113,"children":114},{},[115],{"type":56,"value":116},"Search and list traces; can return large multi-trace payloads",{"type":50,"tag":80,"props":118,"children":119},{},[120,129],{"type":50,"tag":102,"props":121,"children":122},{},[123],{"type":50,"tag":106,"props":124,"children":126},{"className":125},[],[127],{"type":56,"value":128},"posthog:query-llm-trace",{"type":50,"tag":102,"props":130,"children":131},{},[132],{"type":56,"value":133},"Get a single trace by ID with full event tree",{"type":50,"tag":80,"props":135,"children":136},{},[137,146],{"type":50,"tag":102,"props":138,"children":139},{},[140],{"type":50,"tag":106,"props":141,"children":143},{"className":142},[],[144],{"type":56,"value":145},"posthog:read-data-schema",{"type":50,"tag":102,"props":147,"children":148},{},[149],{"type":56,"value":150},"Discover custom event\u002Fperson properties before filtering",{"type":50,"tag":80,"props":152,"children":153},{},[154,163],{"type":50,"tag":102,"props":155,"children":156},{},[157],{"type":50,"tag":106,"props":158,"children":160},{"className":159},[],[161],{"type":56,"value":162},"posthog:execute-sql",{"type":50,"tag":102,"props":164,"children":165},{},[166],{"type":56,"value":167},"Ad-hoc SQL for complex trace analysis",{"type":50,"tag":65,"props":169,"children":171},{"id":170},"event-hierarchy",[172],{"type":56,"value":173},"Event hierarchy",{"type":50,"tag":59,"props":175,"children":176},{},[177,179,186],{"type":56,"value":178},"See the ",{"type":50,"tag":180,"props":181,"children":183},"a",{"href":182},".\u002Freferences\u002Fevents-and-properties.md",[184],{"type":56,"value":185},"event reference",{"type":56,"value":187}," for the full schema.",{"type":50,"tag":189,"props":190,"children":195},"pre",{"className":191,"code":193,"language":56,"meta":194},[192],"language-text","$ai_trace (top-level container)\n  └── $ai_span (logical groupings, e.g. \"RAG retrieval\", \"tool execution\")\n        ├── $ai_generation (individual LLM API call)\n        └── $ai_embedding (embedding creation)\n","",[196],{"type":50,"tag":106,"props":197,"children":198},{"__ignoreMap":194},[199],{"type":56,"value":193},{"type":50,"tag":59,"props":201,"children":202},{},[203,205,211,213,219,221,227],{"type":56,"value":204},"Events are linked via ",{"type":50,"tag":106,"props":206,"children":208},{"className":207},[],[209],{"type":56,"value":210},"$ai_parent_id",{"type":56,"value":212}," → parent's ",{"type":50,"tag":106,"props":214,"children":216},{"className":215},[],[217],{"type":56,"value":218},"$ai_span_id",{"type":56,"value":220}," or ",{"type":50,"tag":106,"props":222,"children":224},{"className":223},[],[225],{"type":56,"value":226},"$ai_trace_id",{"type":56,"value":228},".",{"type":50,"tag":65,"props":230,"children":232},{"id":231},"workflow-debug-a-trace-or-session-from-a-url",[233],{"type":56,"value":234},"Workflow: debug a trace or session from a URL",{"type":50,"tag":236,"props":237,"children":239},"h3",{"id":238},"step-1-classify-the-url",[240],{"type":56,"value":241},"Step 1 — Classify the URL",{"type":50,"tag":59,"props":243,"children":244},{},[245],{"type":56,"value":246},"First inspect the path. Do not treat every UUID-looking value as a trace ID.",{"type":50,"tag":248,"props":249,"children":250},"ul",{},[251,285],{"type":50,"tag":252,"props":253,"children":254},"li",{},[255,261,263,269,271,277,279,284],{"type":50,"tag":106,"props":256,"children":258},{"className":257},[],[259],{"type":56,"value":260},"\u002Fai-observability\u002Ftraces\u002F\u003Ctrace_id>",{"type":56,"value":262}," or legacy ",{"type":50,"tag":106,"props":264,"children":266},{"className":265},[],[267],{"type":56,"value":268},"\u002Fllm-analytics\u002Ftraces\u002F\u003Ctrace_id>",{"type":56,"value":270}," \u002F ",{"type":50,"tag":106,"props":272,"children":274},{"className":273},[],[275],{"type":56,"value":276},"\u002Fllm-observability\u002Ftraces\u002F\u003Ctrace_id>",{"type":56,"value":278}," is a single trace. Fetch it with ",{"type":50,"tag":106,"props":280,"children":282},{"className":281},[],[283],{"type":56,"value":128},{"type":56,"value":228},{"type":50,"tag":252,"props":286,"children":287},{},[288,294,295,301,303,308,310,316],{"type":50,"tag":106,"props":289,"children":291},{"className":290},[],[292],{"type":56,"value":293},"\u002Fai-observability\u002Fsessions\u002F\u003Csession_id>",{"type":56,"value":262},{"type":50,"tag":106,"props":296,"children":298},{"className":297},[],[299],{"type":56,"value":300},"\u002Fllm-analytics\u002Fsessions\u002F\u003Csession_id>",{"type":56,"value":302}," is an AI session, not a trace. Fetch traces with ",{"type":50,"tag":106,"props":304,"children":306},{"className":305},[],[307],{"type":56,"value":111},{"type":56,"value":309}," filtered by event property ",{"type":50,"tag":106,"props":311,"children":313},{"className":312},[],[314],{"type":56,"value":315},"$ai_session_id",{"type":56,"value":228},{"type":50,"tag":59,"props":318,"children":319},{},[320,322,328,329,335,337,343,345,351,353,359,361,366,368,374],{"type":56,"value":321},"Preserve ",{"type":50,"tag":106,"props":323,"children":325},{"className":324},[],[326],{"type":56,"value":327},"date_from",{"type":56,"value":270},{"type":50,"tag":106,"props":330,"children":332},{"className":331},[],[333],{"type":56,"value":334},"date_to",{"type":56,"value":336}," query parameters from the URL when present.\nIf none are present but the URL has a ",{"type":50,"tag":106,"props":338,"children":340},{"className":339},[],[341],{"type":56,"value":342},"timestamp",{"type":56,"value":344}," query parameter, use that timestamp as the anchor and query an absolute window around it, for example ",{"type":50,"tag":106,"props":346,"children":348},{"className":347},[],[349],{"type":56,"value":350},"timestamp - 36h",{"type":56,"value":352}," to ",{"type":50,"tag":106,"props":354,"children":356},{"className":355},[],[357],{"type":56,"value":358},"timestamp + 36h",{"type":56,"value":360},".\nThis handles exact session links whose UI timestamp may be offset from the stored event timestamps while keeping the query bounded.\nIf the URL has neither explicit dates nor ",{"type":50,"tag":106,"props":362,"children":364},{"className":363},[],[365],{"type":56,"value":342},{"type":56,"value":367},", use a safe default like ",{"type":50,"tag":106,"props":369,"children":371},{"className":370},[],[372],{"type":56,"value":373},"{\"date_from\": \"-7d\"}",{"type":56,"value":228},{"type":50,"tag":59,"props":376,"children":377},{},[378,380,386],{"type":56,"value":379},"For exact trace and session URLs, skip schema discovery for the standard ",{"type":50,"tag":106,"props":381,"children":383},{"className":382},[],[384],{"type":56,"value":385},"$ai_*",{"type":56,"value":387}," fields used below. These are AI observability built-ins, not project-specific custom properties.",{"type":50,"tag":236,"props":389,"children":391},{"id":390},"step-2-fetch-trace-data",[392],{"type":56,"value":393},"Step 2 — Fetch trace data",{"type":50,"tag":59,"props":395,"children":396},{},[397,399,404],{"type":56,"value":398},"For a trace URL, call ",{"type":50,"tag":106,"props":400,"children":402},{"className":401},[],[403],{"type":56,"value":128},{"type":56,"value":405}," with:",{"type":50,"tag":189,"props":407,"children":411},{"className":408,"code":409,"language":410,"meta":194,"style":194},"language-json shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","{\n  \"traceId\": \"\u003Ctrace_id>\",\n  \"dateRange\": { \"date_from\": \"-7d\" }\n}\n","json",[412],{"type":50,"tag":106,"props":413,"children":414},{"__ignoreMap":194},[415,427,472,533],{"type":50,"tag":416,"props":417,"children":420},"span",{"class":418,"line":419},"line",1,[421],{"type":50,"tag":416,"props":422,"children":424},{"style":423},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[425],{"type":56,"value":426},"{\n",{"type":50,"tag":416,"props":428,"children":430},{"class":418,"line":429},2,[431,436,442,447,452,457,463,467],{"type":50,"tag":416,"props":432,"children":433},{"style":423},[434],{"type":56,"value":435},"  \"",{"type":50,"tag":416,"props":437,"children":439},{"style":438},"--shiki-light:#9C3EDA;--shiki-default:#C792EA;--shiki-dark:#C792EA",[440],{"type":56,"value":441},"traceId",{"type":50,"tag":416,"props":443,"children":444},{"style":423},[445],{"type":56,"value":446},"\"",{"type":50,"tag":416,"props":448,"children":449},{"style":423},[450],{"type":56,"value":451},":",{"type":50,"tag":416,"props":453,"children":454},{"style":423},[455],{"type":56,"value":456}," \"",{"type":50,"tag":416,"props":458,"children":460},{"style":459},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[461],{"type":56,"value":462},"\u003Ctrace_id>",{"type":50,"tag":416,"props":464,"children":465},{"style":423},[466],{"type":56,"value":446},{"type":50,"tag":416,"props":468,"children":469},{"style":423},[470],{"type":56,"value":471},",\n",{"type":50,"tag":416,"props":473,"children":475},{"class":418,"line":474},3,[476,480,485,489,493,498,502,507,511,515,519,524,528],{"type":50,"tag":416,"props":477,"children":478},{"style":423},[479],{"type":56,"value":435},{"type":50,"tag":416,"props":481,"children":482},{"style":438},[483],{"type":56,"value":484},"dateRange",{"type":50,"tag":416,"props":486,"children":487},{"style":423},[488],{"type":56,"value":446},{"type":50,"tag":416,"props":490,"children":491},{"style":423},[492],{"type":56,"value":451},{"type":50,"tag":416,"props":494,"children":495},{"style":423},[496],{"type":56,"value":497}," {",{"type":50,"tag":416,"props":499,"children":500},{"style":423},[501],{"type":56,"value":456},{"type":50,"tag":416,"props":503,"children":505},{"style":504},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[506],{"type":56,"value":327},{"type":50,"tag":416,"props":508,"children":509},{"style":423},[510],{"type":56,"value":446},{"type":50,"tag":416,"props":512,"children":513},{"style":423},[514],{"type":56,"value":451},{"type":50,"tag":416,"props":516,"children":517},{"style":423},[518],{"type":56,"value":456},{"type":50,"tag":416,"props":520,"children":521},{"style":459},[522],{"type":56,"value":523},"-7d",{"type":50,"tag":416,"props":525,"children":526},{"style":423},[527],{"type":56,"value":446},{"type":50,"tag":416,"props":529,"children":530},{"style":423},[531],{"type":56,"value":532}," }\n",{"type":50,"tag":416,"props":534,"children":536},{"class":418,"line":535},4,[537],{"type":50,"tag":416,"props":538,"children":539},{"style":423},[540],{"type":56,"value":541},"}\n",{"type":50,"tag":59,"props":543,"children":544},{},[545,547,552],{"type":56,"value":546},"For a session URL, call ",{"type":50,"tag":106,"props":548,"children":550},{"className":549},[],[551],{"type":56,"value":111},{"type":56,"value":405},{"type":50,"tag":189,"props":554,"children":556},{"className":408,"code":555,"language":410,"meta":194,"style":194},"{\n  \"dateRange\": { \"date_from\": \"\u003Ctimestamp_minus_36h>\", \"date_to\": \"\u003Ctimestamp_plus_36h>\" },\n  \"filterTestAccounts\": false,\n  \"limit\": 20,\n  \"properties\": [{ \"type\": \"event\", \"key\": \"$ai_session_id\", \"value\": [\"\u003Csession_id>\"], \"operator\": \"exact\" }]\n}\n",[557],{"type":50,"tag":106,"props":558,"children":559},{"__ignoreMap":194},[560,567,658,683,713,881],{"type":50,"tag":416,"props":561,"children":562},{"class":418,"line":419},[563],{"type":50,"tag":416,"props":564,"children":565},{"style":423},[566],{"type":56,"value":426},{"type":50,"tag":416,"props":568,"children":569},{"class":418,"line":429},[570,574,578,582,586,590,594,598,602,606,610,615,619,624,628,632,636,640,644,649,653],{"type":50,"tag":416,"props":571,"children":572},{"style":423},[573],{"type":56,"value":435},{"type":50,"tag":416,"props":575,"children":576},{"style":438},[577],{"type":56,"value":484},{"type":50,"tag":416,"props":579,"children":580},{"style":423},[581],{"type":56,"value":446},{"type":50,"tag":416,"props":583,"children":584},{"style":423},[585],{"type":56,"value":451},{"type":50,"tag":416,"props":587,"children":588},{"style":423},[589],{"type":56,"value":497},{"type":50,"tag":416,"props":591,"children":592},{"style":423},[593],{"type":56,"value":456},{"type":50,"tag":416,"props":595,"children":596},{"style":504},[597],{"type":56,"value":327},{"type":50,"tag":416,"props":599,"children":600},{"style":423},[601],{"type":56,"value":446},{"type":50,"tag":416,"props":603,"children":604},{"style":423},[605],{"type":56,"value":451},{"type":50,"tag":416,"props":607,"children":608},{"style":423},[609],{"type":56,"value":456},{"type":50,"tag":416,"props":611,"children":612},{"style":459},[613],{"type":56,"value":614},"\u003Ctimestamp_minus_36h>",{"type":50,"tag":416,"props":616,"children":617},{"style":423},[618],{"type":56,"value":446},{"type":50,"tag":416,"props":620,"children":621},{"style":423},[622],{"type":56,"value":623},",",{"type":50,"tag":416,"props":625,"children":626},{"style":423},[627],{"type":56,"value":456},{"type":50,"tag":416,"props":629,"children":630},{"style":504},[631],{"type":56,"value":334},{"type":50,"tag":416,"props":633,"children":634},{"style":423},[635],{"type":56,"value":446},{"type":50,"tag":416,"props":637,"children":638},{"style":423},[639],{"type":56,"value":451},{"type":50,"tag":416,"props":641,"children":642},{"style":423},[643],{"type":56,"value":456},{"type":50,"tag":416,"props":645,"children":646},{"style":459},[647],{"type":56,"value":648},"\u003Ctimestamp_plus_36h>",{"type":50,"tag":416,"props":650,"children":651},{"style":423},[652],{"type":56,"value":446},{"type":50,"tag":416,"props":654,"children":655},{"style":423},[656],{"type":56,"value":657}," },\n",{"type":50,"tag":416,"props":659,"children":660},{"class":418,"line":474},[661,665,670,674,678],{"type":50,"tag":416,"props":662,"children":663},{"style":423},[664],{"type":56,"value":435},{"type":50,"tag":416,"props":666,"children":667},{"style":438},[668],{"type":56,"value":669},"filterTestAccounts",{"type":50,"tag":416,"props":671,"children":672},{"style":423},[673],{"type":56,"value":446},{"type":50,"tag":416,"props":675,"children":676},{"style":423},[677],{"type":56,"value":451},{"type":50,"tag":416,"props":679,"children":680},{"style":423},[681],{"type":56,"value":682}," false,\n",{"type":50,"tag":416,"props":684,"children":685},{"class":418,"line":535},[686,690,695,699,703,709],{"type":50,"tag":416,"props":687,"children":688},{"style":423},[689],{"type":56,"value":435},{"type":50,"tag":416,"props":691,"children":692},{"style":438},[693],{"type":56,"value":694},"limit",{"type":50,"tag":416,"props":696,"children":697},{"style":423},[698],{"type":56,"value":446},{"type":50,"tag":416,"props":700,"children":701},{"style":423},[702],{"type":56,"value":451},{"type":50,"tag":416,"props":704,"children":706},{"style":705},"--shiki-light:#F76D47;--shiki-default:#F78C6C;--shiki-dark:#F78C6C",[707],{"type":56,"value":708}," 20",{"type":50,"tag":416,"props":710,"children":711},{"style":423},[712],{"type":56,"value":471},{"type":50,"tag":416,"props":714,"children":716},{"class":418,"line":715},5,[717,721,726,730,734,739,743,748,752,756,760,765,769,773,777,782,786,790,794,798,802,806,810,815,819,823,828,832,837,841,846,850,855,859,863,867,872,876],{"type":50,"tag":416,"props":718,"children":719},{"style":423},[720],{"type":56,"value":435},{"type":50,"tag":416,"props":722,"children":723},{"style":438},[724],{"type":56,"value":725},"properties",{"type":50,"tag":416,"props":727,"children":728},{"style":423},[729],{"type":56,"value":446},{"type":50,"tag":416,"props":731,"children":732},{"style":423},[733],{"type":56,"value":451},{"type":50,"tag":416,"props":735,"children":736},{"style":423},[737],{"type":56,"value":738}," [{",{"type":50,"tag":416,"props":740,"children":741},{"style":423},[742],{"type":56,"value":456},{"type":50,"tag":416,"props":744,"children":745},{"style":504},[746],{"type":56,"value":747},"type",{"type":50,"tag":416,"props":749,"children":750},{"style":423},[751],{"type":56,"value":446},{"type":50,"tag":416,"props":753,"children":754},{"style":423},[755],{"type":56,"value":451},{"type":50,"tag":416,"props":757,"children":758},{"style":423},[759],{"type":56,"value":456},{"type":50,"tag":416,"props":761,"children":762},{"style":459},[763],{"type":56,"value":764},"event",{"type":50,"tag":416,"props":766,"children":767},{"style":423},[768],{"type":56,"value":446},{"type":50,"tag":416,"props":770,"children":771},{"style":423},[772],{"type":56,"value":623},{"type":50,"tag":416,"props":774,"children":775},{"style":423},[776],{"type":56,"value":456},{"type":50,"tag":416,"props":778,"children":779},{"style":504},[780],{"type":56,"value":781},"key",{"type":50,"tag":416,"props":783,"children":784},{"style":423},[785],{"type":56,"value":446},{"type":50,"tag":416,"props":787,"children":788},{"style":423},[789],{"type":56,"value":451},{"type":50,"tag":416,"props":791,"children":792},{"style":423},[793],{"type":56,"value":456},{"type":50,"tag":416,"props":795,"children":796},{"style":459},[797],{"type":56,"value":315},{"type":50,"tag":416,"props":799,"children":800},{"style":423},[801],{"type":56,"value":446},{"type":50,"tag":416,"props":803,"children":804},{"style":423},[805],{"type":56,"value":623},{"type":50,"tag":416,"props":807,"children":808},{"style":423},[809],{"type":56,"value":456},{"type":50,"tag":416,"props":811,"children":812},{"style":504},[813],{"type":56,"value":814},"value",{"type":50,"tag":416,"props":816,"children":817},{"style":423},[818],{"type":56,"value":446},{"type":50,"tag":416,"props":820,"children":821},{"style":423},[822],{"type":56,"value":451},{"type":50,"tag":416,"props":824,"children":825},{"style":423},[826],{"type":56,"value":827}," [",{"type":50,"tag":416,"props":829,"children":830},{"style":423},[831],{"type":56,"value":446},{"type":50,"tag":416,"props":833,"children":834},{"style":459},[835],{"type":56,"value":836},"\u003Csession_id>",{"type":50,"tag":416,"props":838,"children":839},{"style":423},[840],{"type":56,"value":446},{"type":50,"tag":416,"props":842,"children":843},{"style":423},[844],{"type":56,"value":845},"],",{"type":50,"tag":416,"props":847,"children":848},{"style":423},[849],{"type":56,"value":456},{"type":50,"tag":416,"props":851,"children":852},{"style":504},[853],{"type":56,"value":854},"operator",{"type":50,"tag":416,"props":856,"children":857},{"style":423},[858],{"type":56,"value":446},{"type":50,"tag":416,"props":860,"children":861},{"style":423},[862],{"type":56,"value":451},{"type":50,"tag":416,"props":864,"children":865},{"style":423},[866],{"type":56,"value":456},{"type":50,"tag":416,"props":868,"children":869},{"style":459},[870],{"type":56,"value":871},"exact",{"type":50,"tag":416,"props":873,"children":874},{"style":423},[875],{"type":56,"value":446},{"type":50,"tag":416,"props":877,"children":878},{"style":423},[879],{"type":56,"value":880}," }]\n",{"type":50,"tag":416,"props":882,"children":884},{"class":418,"line":883},6,[885],{"type":50,"tag":416,"props":886,"children":887},{"style":423},[888],{"type":56,"value":541},{"type":50,"tag":59,"props":890,"children":891},{},[892,894,899,900,905,907,912,914,920,922,928],{"type":56,"value":893},"Use the URL's ",{"type":50,"tag":106,"props":895,"children":897},{"className":896},[],[898],{"type":56,"value":327},{"type":56,"value":270},{"type":50,"tag":106,"props":901,"children":903},{"className":902},[],[904],{"type":56,"value":334},{"type":56,"value":906}," values in the session query if present.\nIf the URL only has ",{"type":50,"tag":106,"props":908,"children":910},{"className":909},[],[911],{"type":56,"value":342},{"type":56,"value":913},", calculate the absolute date range from that timestamp instead of using a relative range like ",{"type":50,"tag":106,"props":915,"children":917},{"className":916},[],[918],{"type":56,"value":919},"-1h",{"type":56,"value":921},".\nSet ",{"type":50,"tag":106,"props":923,"children":925},{"className":924},[],[926],{"type":56,"value":927},"filterTestAccounts: false",{"type":56,"value":929}," for an exact URL so the requested trace is not hidden by account filters.",{"type":50,"tag":59,"props":931,"children":932},{},[933],{"type":56,"value":934},"The result contains the event tree with all properties.\nThe response may be large — when it exceeds the inline limit, Claude Code auto-persists it to a file.",{"type":50,"tag":59,"props":936,"children":937},{},[938],{"type":56,"value":939},"From the result you get:",{"type":50,"tag":248,"props":941,"children":942},{},[943,964,977,982,992],{"type":50,"tag":252,"props":944,"children":945},{},[946,948,954,956,962],{"type":56,"value":947},"Every event with its type (",{"type":50,"tag":106,"props":949,"children":951},{"className":950},[],[952],{"type":56,"value":953},"$ai_span",{"type":56,"value":955},", ",{"type":50,"tag":106,"props":957,"children":959},{"className":958},[],[960],{"type":56,"value":961},"$ai_generation",{"type":56,"value":963},", etc.)",{"type":50,"tag":252,"props":965,"children":966},{},[967,969,975],{"type":56,"value":968},"Span names (",{"type":50,"tag":106,"props":970,"children":972},{"className":971},[],[973],{"type":56,"value":974},"$ai_span_name",{"type":56,"value":976},") — these are the tool\u002Fstep names",{"type":50,"tag":252,"props":978,"children":979},{},[980],{"type":56,"value":981},"Latency, error flags, models used",{"type":50,"tag":252,"props":983,"children":984},{},[985,987],{"type":56,"value":986},"Parent-child relationships via ",{"type":50,"tag":106,"props":988,"children":990},{"className":989},[],[991],{"type":56,"value":210},{"type":50,"tag":252,"props":993,"children":994},{},[995,1001,1003,1009],{"type":50,"tag":106,"props":996,"children":998},{"className":997},[],[999],{"type":56,"value":1000},"_posthogUrl",{"type":56,"value":1002}," — ",{"type":50,"tag":1004,"props":1005,"children":1006},"strong",{},[1007],{"type":56,"value":1008},"always include this in your response",{"type":56,"value":1010}," so the user can click through to the UI",{"type":50,"tag":236,"props":1012,"children":1014},{"id":1013},"step-3-parse-large-results-with-scripts",[1015],{"type":56,"value":1016},"Step 3 — Parse large results with scripts",{"type":50,"tag":59,"props":1018,"children":1019},{},[1020,1022,1028,1030,1036,1038,1044],{"type":56,"value":1021},"When the result is persisted to a file (large traces with full ",{"type":50,"tag":106,"props":1023,"children":1025},{"className":1024},[],[1026],{"type":56,"value":1027},"$ai_input",{"type":56,"value":1029},"\u002F",{"type":50,"tag":106,"props":1031,"children":1033},{"className":1032},[],[1034],{"type":56,"value":1035},"$ai_output_choices",{"type":56,"value":1037},"),\nuse the ",{"type":50,"tag":180,"props":1039,"children":1041},{"href":1040},".\u002Fscripts\u002F",[1042],{"type":56,"value":1043},"parsing scripts",{"type":56,"value":1045}," to explore it.",{"type":50,"tag":59,"props":1047,"children":1048},{},[1049,1054],{"type":50,"tag":1004,"props":1050,"children":1051},{},[1052],{"type":56,"value":1053},"Start with the summary",{"type":56,"value":1055}," to get the full picture, then drill into specifics:",{"type":50,"tag":189,"props":1057,"children":1061},{"className":1058,"code":1059,"language":1060,"meta":194,"style":194},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","# 1. Overview: metadata, tool calls, final output, errors\npython3 scripts\u002Fprint_summary.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 2. Timeline: chronological event list with truncated I\u002FO\npython3 scripts\u002Fprint_timeline.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 3. Drill into a specific span's full input\u002Foutput\nSPAN=\"tool_name\" python3 scripts\u002Fextract_span.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 4. Full conversation with thinking blocks and tool calls\npython3 scripts\u002Fextract_conversation.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n\n# 5. Search for a keyword across all properties\nSEARCH=\"keyword\" python3 scripts\u002Fsearch_traces.py \u002Fpath\u002Fto\u002Fpersisted-file.json\n","bash",[1062],{"type":50,"tag":106,"props":1063,"children":1064},{"__ignoreMap":194},[1065,1074,1092,1101,1109,1125,1132,1141,1183,1191,1200,1216,1224,1233],{"type":50,"tag":416,"props":1066,"children":1067},{"class":418,"line":419},[1068],{"type":50,"tag":416,"props":1069,"children":1071},{"style":1070},"--shiki-light:#90A4AE;--shiki-light-font-style:italic;--shiki-default:#546E7A;--shiki-default-font-style:italic;--shiki-dark:#676E95;--shiki-dark-font-style:italic",[1072],{"type":56,"value":1073},"# 1. Overview: metadata, tool calls, final output, errors\n",{"type":50,"tag":416,"props":1075,"children":1076},{"class":418,"line":429},[1077,1082,1087],{"type":50,"tag":416,"props":1078,"children":1079},{"style":504},[1080],{"type":56,"value":1081},"python3",{"type":50,"tag":416,"props":1083,"children":1084},{"style":459},[1085],{"type":56,"value":1086}," scripts\u002Fprint_summary.py",{"type":50,"tag":416,"props":1088,"children":1089},{"style":459},[1090],{"type":56,"value":1091}," \u002Fpath\u002Fto\u002Fpersisted-file.json\n",{"type":50,"tag":416,"props":1093,"children":1094},{"class":418,"line":474},[1095],{"type":50,"tag":416,"props":1096,"children":1098},{"emptyLinePlaceholder":1097},true,[1099],{"type":56,"value":1100},"\n",{"type":50,"tag":416,"props":1102,"children":1103},{"class":418,"line":535},[1104],{"type":50,"tag":416,"props":1105,"children":1106},{"style":1070},[1107],{"type":56,"value":1108},"# 2. Timeline: chronological event list with truncated I\u002FO\n",{"type":50,"tag":416,"props":1110,"children":1111},{"class":418,"line":715},[1112,1116,1121],{"type":50,"tag":416,"props":1113,"children":1114},{"style":504},[1115],{"type":56,"value":1081},{"type":50,"tag":416,"props":1117,"children":1118},{"style":459},[1119],{"type":56,"value":1120}," scripts\u002Fprint_timeline.py",{"type":50,"tag":416,"props":1122,"children":1123},{"style":459},[1124],{"type":56,"value":1091},{"type":50,"tag":416,"props":1126,"children":1127},{"class":418,"line":883},[1128],{"type":50,"tag":416,"props":1129,"children":1130},{"emptyLinePlaceholder":1097},[1131],{"type":56,"value":1100},{"type":50,"tag":416,"props":1133,"children":1135},{"class":418,"line":1134},7,[1136],{"type":50,"tag":416,"props":1137,"children":1138},{"style":1070},[1139],{"type":56,"value":1140},"# 3. Drill into a specific span's full input\u002Foutput\n",{"type":50,"tag":416,"props":1142,"children":1144},{"class":418,"line":1143},8,[1145,1151,1156,1160,1165,1169,1174,1179],{"type":50,"tag":416,"props":1146,"children":1148},{"style":1147},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[1149],{"type":56,"value":1150},"SPAN",{"type":50,"tag":416,"props":1152,"children":1153},{"style":423},[1154],{"type":56,"value":1155},"=",{"type":50,"tag":416,"props":1157,"children":1158},{"style":423},[1159],{"type":56,"value":446},{"type":50,"tag":416,"props":1161,"children":1162},{"style":459},[1163],{"type":56,"value":1164},"tool_name",{"type":50,"tag":416,"props":1166,"children":1167},{"style":423},[1168],{"type":56,"value":446},{"type":50,"tag":416,"props":1170,"children":1171},{"style":504},[1172],{"type":56,"value":1173}," python3",{"type":50,"tag":416,"props":1175,"children":1176},{"style":459},[1177],{"type":56,"value":1178}," scripts\u002Fextract_span.py",{"type":50,"tag":416,"props":1180,"children":1181},{"style":459},[1182],{"type":56,"value":1091},{"type":50,"tag":416,"props":1184,"children":1186},{"class":418,"line":1185},9,[1187],{"type":50,"tag":416,"props":1188,"children":1189},{"emptyLinePlaceholder":1097},[1190],{"type":56,"value":1100},{"type":50,"tag":416,"props":1192,"children":1194},{"class":418,"line":1193},10,[1195],{"type":50,"tag":416,"props":1196,"children":1197},{"style":1070},[1198],{"type":56,"value":1199},"# 4. Full conversation with thinking blocks and tool calls\n",{"type":50,"tag":416,"props":1201,"children":1202},{"class":418,"line":33},[1203,1207,1212],{"type":50,"tag":416,"props":1204,"children":1205},{"style":504},[1206],{"type":56,"value":1081},{"type":50,"tag":416,"props":1208,"children":1209},{"style":459},[1210],{"type":56,"value":1211}," scripts\u002Fextract_conversation.py",{"type":50,"tag":416,"props":1213,"children":1214},{"style":459},[1215],{"type":56,"value":1091},{"type":50,"tag":416,"props":1217,"children":1219},{"class":418,"line":1218},12,[1220],{"type":50,"tag":416,"props":1221,"children":1222},{"emptyLinePlaceholder":1097},[1223],{"type":56,"value":1100},{"type":50,"tag":416,"props":1225,"children":1227},{"class":418,"line":1226},13,[1228],{"type":50,"tag":416,"props":1229,"children":1230},{"style":1070},[1231],{"type":56,"value":1232},"# 5. Search for a keyword across all properties\n",{"type":50,"tag":416,"props":1234,"children":1236},{"class":418,"line":1235},14,[1237,1242,1246,1250,1255,1259,1263,1268],{"type":50,"tag":416,"props":1238,"children":1239},{"style":1147},[1240],{"type":56,"value":1241},"SEARCH",{"type":50,"tag":416,"props":1243,"children":1244},{"style":423},[1245],{"type":56,"value":1155},{"type":50,"tag":416,"props":1247,"children":1248},{"style":423},[1249],{"type":56,"value":446},{"type":50,"tag":416,"props":1251,"children":1252},{"style":459},[1253],{"type":56,"value":1254},"keyword",{"type":50,"tag":416,"props":1256,"children":1257},{"style":423},[1258],{"type":56,"value":446},{"type":50,"tag":416,"props":1260,"children":1261},{"style":504},[1262],{"type":56,"value":1173},{"type":50,"tag":416,"props":1264,"children":1265},{"style":459},[1266],{"type":56,"value":1267}," scripts\u002Fsearch_traces.py",{"type":50,"tag":416,"props":1269,"children":1270},{"style":459},[1271],{"type":56,"value":1091},{"type":50,"tag":59,"props":1273,"children":1274},{},[1275,1277,1283],{"type":56,"value":1276},"All scripts support ",{"type":50,"tag":106,"props":1278,"children":1280},{"className":1279},[],[1281],{"type":56,"value":1282},"MAX_LEN=N",{"type":56,"value":1284}," env var to control truncation (0 = unlimited).",{"type":50,"tag":65,"props":1286,"children":1288},{"id":1287},"investigation-patterns",[1289],{"type":56,"value":1290},"Investigation patterns",{"type":50,"tag":236,"props":1292,"children":1294},{"id":1293},"did-the-agent-use-the-tool-correctly",[1295],{"type":56,"value":1296},"\"Did the agent use the tool correctly?\"",{"type":50,"tag":1298,"props":1299,"children":1300},"ol",{},[1301,1320,1333,1345],{"type":50,"tag":252,"props":1302,"children":1303},{},[1304,1306,1311,1313,1318],{"type":56,"value":1305},"Find the ",{"type":50,"tag":106,"props":1307,"children":1309},{"className":1308},[],[1310],{"type":56,"value":953},{"type":56,"value":1312}," for the tool call (look at ",{"type":50,"tag":106,"props":1314,"children":1316},{"className":1315},[],[1317],{"type":56,"value":974},{"type":56,"value":1319},")",{"type":50,"tag":252,"props":1321,"children":1322},{},[1323,1325,1331],{"type":56,"value":1324},"Check ",{"type":50,"tag":106,"props":1326,"children":1328},{"className":1327},[],[1329],{"type":56,"value":1330},"$ai_input_state",{"type":56,"value":1332}," — what arguments were passed to the tool?",{"type":50,"tag":252,"props":1334,"children":1335},{},[1336,1337,1343],{"type":56,"value":1324},{"type":50,"tag":106,"props":1338,"children":1340},{"className":1339},[],[1341],{"type":56,"value":1342},"$ai_output_state",{"type":56,"value":1344}," — what did the tool return?",{"type":50,"tag":252,"props":1346,"children":1347},{},[1348,1349,1355],{"type":56,"value":1324},{"type":50,"tag":106,"props":1350,"children":1352},{"className":1351},[],[1353],{"type":56,"value":1354},"$ai_is_error",{"type":56,"value":1356}," — did the tool call fail?",{"type":50,"tag":236,"props":1358,"children":1360},{"id":1359},"was-the-context-correct-were-the-right-files-surfaced",[1361],{"type":56,"value":1362},"\"Was the context correct?\" \u002F \"Were the right files surfaced?\"",{"type":50,"tag":1298,"props":1364,"children":1365},{},[1366,1377,1388,1400],{"type":50,"tag":252,"props":1367,"children":1368},{},[1369,1370,1375],{"type":56,"value":1305},{"type":50,"tag":106,"props":1371,"children":1373},{"className":1372},[],[1374],{"type":56,"value":961},{"type":56,"value":1376}," event where the LLM made the decision",{"type":50,"tag":252,"props":1378,"children":1379},{},[1380,1381,1386],{"type":56,"value":1324},{"type":50,"tag":106,"props":1382,"children":1384},{"className":1383},[],[1385],{"type":56,"value":1027},{"type":56,"value":1387}," — this is the full message history the LLM saw",{"type":50,"tag":252,"props":1389,"children":1390},{},[1391,1393,1398],{"type":56,"value":1392},"Look at preceding ",{"type":50,"tag":106,"props":1394,"children":1396},{"className":1395},[],[1397],{"type":56,"value":953},{"type":56,"value":1399}," events for retrieval\u002Fsearch steps",{"type":50,"tag":252,"props":1401,"children":1402},{},[1403,1405,1410],{"type":56,"value":1404},"Check their ",{"type":50,"tag":106,"props":1406,"children":1408},{"className":1407},[],[1409],{"type":56,"value":1342},{"type":56,"value":1411}," — what content was retrieved and fed to the LLM?",{"type":50,"tag":236,"props":1413,"children":1415},{"id":1414},"did-the-subagent-work",[1416],{"type":56,"value":1417},"\"Did the subagent work?\"",{"type":50,"tag":1298,"props":1419,"children":1420},{},[1421,1432,1437,1454],{"type":50,"tag":252,"props":1422,"children":1423},{},[1424,1426,1431],{"type":56,"value":1425},"In the structural overview, find spans that are children of other spans (via ",{"type":50,"tag":106,"props":1427,"children":1429},{"className":1428},[],[1430],{"type":56,"value":210},{"type":56,"value":1319},{"type":50,"tag":252,"props":1433,"children":1434},{},[1435],{"type":56,"value":1436},"The parent span is the orchestrator; child spans are subagent steps",{"type":50,"tag":252,"props":1438,"children":1439},{},[1440,1442,1447,1449],{"type":56,"value":1441},"Check each child's ",{"type":50,"tag":106,"props":1443,"children":1445},{"className":1444},[],[1446],{"type":56,"value":1342},{"type":56,"value":1448}," and ",{"type":50,"tag":106,"props":1450,"children":1452},{"className":1451},[],[1453],{"type":56,"value":1354},{"type":50,"tag":252,"props":1455,"children":1456},{},[1457,1459,1464],{"type":56,"value":1458},"If a child span contains ",{"type":50,"tag":106,"props":1460,"children":1462},{"className":1461},[],[1463],{"type":56,"value":961},{"type":56,"value":1465}," events, those are the subagent's LLM calls",{"type":50,"tag":236,"props":1467,"children":1469},{"id":1468},"why-did-the-llm-say-x",[1470],{"type":56,"value":1471},"\"Why did the LLM say X?\"",{"type":50,"tag":1298,"props":1473,"children":1474},{},[1475,1494,1499],{"type":50,"tag":252,"props":1476,"children":1477},{},[1478,1480,1486,1488],{"type":56,"value":1479},"Use ",{"type":50,"tag":106,"props":1481,"children":1483},{"className":1482},[],[1484],{"type":56,"value":1485},"search_traces.py",{"type":56,"value":1487}," to find where the text appears: ",{"type":50,"tag":106,"props":1489,"children":1491},{"className":1490},[],[1492],{"type":56,"value":1493},"SEARCH=\"the text\" python3 scripts\u002Fsearch_traces.py FILE",{"type":50,"tag":252,"props":1495,"children":1496},{},[1497],{"type":56,"value":1498},"This shows which event and property path contains it",{"type":50,"tag":252,"props":1500,"children":1501},{},[1502,1504,1509],{"type":56,"value":1503},"Check the ",{"type":50,"tag":106,"props":1505,"children":1507},{"className":1506},[],[1508],{"type":56,"value":1027},{"type":56,"value":1510}," of that generation to see what the LLM was told before it said X",{"type":50,"tag":65,"props":1512,"children":1514},{"id":1513},"constructing-ui-links",[1515],{"type":56,"value":1516},"Constructing UI links",{"type":50,"tag":59,"props":1518,"children":1519},{},[1520,1522,1527],{"type":56,"value":1521},"The trace tools return ",{"type":50,"tag":106,"props":1523,"children":1525},{"className":1524},[],[1526],{"type":56,"value":1000},{"type":56,"value":1528}," — always surface this to the user.",{"type":50,"tag":59,"props":1530,"children":1531},{},[1532],{"type":56,"value":1533},"You can also construct links manually:",{"type":50,"tag":248,"props":1535,"children":1536},{},[1537,1553],{"type":50,"tag":252,"props":1538,"children":1539},{},[1540,1545,1547],{"type":50,"tag":1004,"props":1541,"children":1542},{},[1543],{"type":56,"value":1544},"Trace detail",{"type":56,"value":1546},": ",{"type":50,"tag":106,"props":1548,"children":1550},{"className":1549},[],[1551],{"type":56,"value":1552},"https:\u002F\u002Fapp.posthog.com\u002Fai-observability\u002Ftraces\u002F\u003Ctrace_id>?timestamp=\u003Curl_encoded_timestamp>&event=\u003Coptional_event_id>",{"type":50,"tag":252,"props":1554,"children":1555},{},[1556,1561,1563,1568,1570],{"type":50,"tag":1004,"props":1557,"children":1558},{},[1559],{"type":56,"value":1560},"Traces list with filters",{"type":56,"value":1562},": returned in ",{"type":50,"tag":106,"props":1564,"children":1566},{"className":1565},[],[1567],{"type":56,"value":1000},{"type":56,"value":1569}," from ",{"type":50,"tag":106,"props":1571,"children":1573},{"className":1572},[],[1574],{"type":56,"value":1575},"query-llm-traces-list",{"type":50,"tag":59,"props":1577,"children":1578},{},[1579,1581,1586,1588,1593,1595,1601,1603,1609],{"type":56,"value":1580},"The ",{"type":50,"tag":106,"props":1582,"children":1584},{"className":1583},[],[1585],{"type":56,"value":342},{"type":56,"value":1587}," query param is ",{"type":50,"tag":1004,"props":1589,"children":1590},{},[1591],{"type":56,"value":1592},"required",{"type":56,"value":1594}," — use the ",{"type":50,"tag":106,"props":1596,"children":1598},{"className":1597},[],[1599],{"type":56,"value":1600},"createdAt",{"type":56,"value":1602}," of the earliest event in the trace, URL-encoded (e.g. ",{"type":50,"tag":106,"props":1604,"children":1606},{"className":1605},[],[1607],{"type":56,"value":1608},"timestamp=2026-04-01T19%3A39%3A20Z",{"type":56,"value":1610},").",{"type":50,"tag":59,"props":1612,"children":1613},{},[1614],{"type":56,"value":1615},"When presenting findings, always include the relevant PostHog URL so the user can verify.",{"type":50,"tag":65,"props":1617,"children":1619},{"id":1618},"finding-traces",[1620],{"type":56,"value":1621},"Finding traces",{"type":50,"tag":59,"props":1623,"children":1624},{},[1625,1626,1631],{"type":56,"value":1479},{"type":50,"tag":106,"props":1627,"children":1629},{"className":1628},[],[1630],{"type":56,"value":111},{"type":56,"value":1632}," to search and filter traces.",{"type":50,"tag":59,"props":1634,"children":1635},{},[1636,1641,1643,1648],{"type":50,"tag":1004,"props":1637,"children":1638},{},[1639],{"type":56,"value":1640},"CRITICAL: Never assume event names, property names, or property values from training data.",{"type":56,"value":1642},"\nEvery project instruments different custom properties. For open-ended searches and custom filters, call\n",{"type":50,"tag":106,"props":1644,"children":1646},{"className":1645},[],[1647],{"type":56,"value":145},{"type":56,"value":1649}," first to discover what properties and values actually exist in the project's\ndata before constructing filters.",{"type":50,"tag":59,"props":1651,"children":1652},{},[1653,1655,1660,1661,1666],{"type":56,"value":1654},"The exception is exact AI observability trace\u002Fsession URLs: use the built-in ",{"type":50,"tag":106,"props":1656,"children":1658},{"className":1657},[],[1659],{"type":56,"value":226},{"type":56,"value":270},{"type":50,"tag":106,"props":1662,"children":1664},{"className":1663},[],[1665],{"type":56,"value":315},{"type":56,"value":1667},"\nfields directly and skip schema discovery.",{"type":50,"tag":236,"props":1669,"children":1671},{"id":1670},"discovering-the-schema-first",[1672],{"type":56,"value":1673},"Discovering the schema first",{"type":50,"tag":59,"props":1675,"children":1676},{},[1677],{"type":56,"value":1678},"Before filtering traces, discover what's available:",{"type":50,"tag":1298,"props":1680,"children":1681},{},[1682,1714,1744],{"type":50,"tag":252,"props":1683,"children":1684},{},[1685,1690,1692,1697,1699,1705,1707,1712],{"type":50,"tag":1004,"props":1686,"children":1687},{},[1688],{"type":56,"value":1689},"Confirm AI events exist",{"type":56,"value":1691}," — call ",{"type":50,"tag":106,"props":1693,"children":1695},{"className":1694},[],[1696],{"type":56,"value":145},{"type":56,"value":1698}," with ",{"type":50,"tag":106,"props":1700,"children":1702},{"className":1701},[],[1703],{"type":56,"value":1704},"kind: \"events\"",{"type":56,"value":1706}," and look for ",{"type":50,"tag":106,"props":1708,"children":1710},{"className":1709},[],[1711],{"type":56,"value":385},{"type":56,"value":1713}," events",{"type":50,"tag":252,"props":1715,"children":1716},{},[1717,1722,1723,1728,1729,1735,1736,1742],{"type":50,"tag":1004,"props":1718,"children":1719},{},[1720],{"type":56,"value":1721},"Find filterable properties",{"type":56,"value":1691},{"type":50,"tag":106,"props":1724,"children":1726},{"className":1725},[],[1727],{"type":56,"value":145},{"type":56,"value":1698},{"type":50,"tag":106,"props":1730,"children":1732},{"className":1731},[],[1733],{"type":56,"value":1734},"kind: \"event_properties\"",{"type":56,"value":1448},{"type":50,"tag":106,"props":1737,"children":1739},{"className":1738},[],[1740],{"type":56,"value":1741},"event_name: \"$ai_generation\"",{"type":56,"value":1743}," (or another AI event) to see what properties are captured",{"type":50,"tag":252,"props":1745,"children":1746},{},[1747,1752,1753,1758,1759,1765,1766,1771,1773,1779],{"type":50,"tag":1004,"props":1748,"children":1749},{},[1750],{"type":56,"value":1751},"Get actual values",{"type":56,"value":1691},{"type":50,"tag":106,"props":1754,"children":1756},{"className":1755},[],[1757],{"type":56,"value":145},{"type":56,"value":1698},{"type":50,"tag":106,"props":1760,"children":1762},{"className":1761},[],[1763],{"type":56,"value":1764},"kind: \"event_property_values\"",{"type":56,"value":955},{"type":50,"tag":106,"props":1767,"children":1769},{"className":1768},[],[1770],{"type":56,"value":1741},{"type":56,"value":1772},", and ",{"type":50,"tag":106,"props":1774,"children":1776},{"className":1775},[],[1777],{"type":56,"value":1778},"property_name: \"$ai_model\"",{"type":56,"value":1780}," to see real model names in use",{"type":50,"tag":59,"props":1782,"children":1783},{},[1784,1786,1791],{"type":56,"value":1785},"Only then construct the ",{"type":50,"tag":106,"props":1787,"children":1789},{"className":1788},[],[1790],{"type":56,"value":1575},{"type":56,"value":1792}," call with property filters.",{"type":50,"tag":59,"props":1794,"children":1795},{},[1796,1798,1804,1805,1811,1812,1818],{"type":56,"value":1797},"This is especially important for custom properties like ",{"type":50,"tag":106,"props":1799,"children":1801},{"className":1800},[],[1802],{"type":56,"value":1803},"project_id",{"type":56,"value":955},{"type":50,"tag":106,"props":1806,"children":1808},{"className":1807},[],[1809],{"type":56,"value":1810},"conversation_id",{"type":56,"value":955},{"type":50,"tag":106,"props":1813,"children":1815},{"className":1814},[],[1816],{"type":56,"value":1817},"user_tier",{"type":56,"value":1819},", etc. — these vary per project and cannot be guessed.",{"type":50,"tag":59,"props":1821,"children":1822},{},[1823,1825,1830,1832,1838],{"type":56,"value":1824},"Do not confirm ",{"type":50,"tag":106,"props":1826,"children":1828},{"className":1827},[],[1829],{"type":56,"value":385},{"type":56,"value":1831}," properties, but confirm any other like ",{"type":50,"tag":106,"props":1833,"children":1835},{"className":1834},[],[1836],{"type":56,"value":1837},"email",{"type":56,"value":1839}," of a person.",{"type":50,"tag":236,"props":1841,"children":1843},{"id":1842},"by-filters",[1844],{"type":56,"value":1845},"By filters",{"type":50,"tag":189,"props":1847,"children":1849},{"className":408,"code":1848,"language":410,"meta":194,"style":194},"posthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"limit\": 20,\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"$ai_model\", \"value\": \"gpt-4o\", \"operator\": \"exact\"}\n  ]\n}\n",[1850],{"type":50,"tag":106,"props":1851,"children":1852},{"__ignoreMap":194},[1853,1861,1868,1924,1948,1975,1999,2137,2145],{"type":50,"tag":416,"props":1854,"children":1855},{"class":418,"line":419},[1856],{"type":50,"tag":416,"props":1857,"children":1858},{"style":1147},[1859],{"type":56,"value":1860},"posthog:query-llm-traces-list\n",{"type":50,"tag":416,"props":1862,"children":1863},{"class":418,"line":429},[1864],{"type":50,"tag":416,"props":1865,"children":1866},{"style":423},[1867],{"type":56,"value":426},{"type":50,"tag":416,"props":1869,"children":1870},{"class":418,"line":474},[1871,1875,1879,1883,1887,1891,1895,1899,1903,1907,1911,1915,1919],{"type":50,"tag":416,"props":1872,"children":1873},{"style":423},[1874],{"type":56,"value":435},{"type":50,"tag":416,"props":1876,"children":1877},{"style":438},[1878],{"type":56,"value":484},{"type":50,"tag":416,"props":1880,"children":1881},{"style":423},[1882],{"type":56,"value":446},{"type":50,"tag":416,"props":1884,"children":1885},{"style":423},[1886],{"type":56,"value":451},{"type":50,"tag":416,"props":1888,"children":1889},{"style":423},[1890],{"type":56,"value":497},{"type":50,"tag":416,"props":1892,"children":1893},{"style":423},[1894],{"type":56,"value":446},{"type":50,"tag":416,"props":1896,"children":1897},{"style":504},[1898],{"type":56,"value":327},{"type":50,"tag":416,"props":1900,"children":1901},{"style":423},[1902],{"type":56,"value":446},{"type":50,"tag":416,"props":1904,"children":1905},{"style":423},[1906],{"type":56,"value":451},{"type":50,"tag":416,"props":1908,"children":1909},{"style":423},[1910],{"type":56,"value":456},{"type":50,"tag":416,"props":1912,"children":1913},{"style":459},[1914],{"type":56,"value":919},{"type":50,"tag":416,"props":1916,"children":1917},{"style":423},[1918],{"type":56,"value":446},{"type":50,"tag":416,"props":1920,"children":1921},{"style":423},[1922],{"type":56,"value":1923},"},\n",{"type":50,"tag":416,"props":1925,"children":1926},{"class":418,"line":535},[1927,1931,1935,1939,1943],{"type":50,"tag":416,"props":1928,"children":1929},{"style":423},[1930],{"type":56,"value":435},{"type":50,"tag":416,"props":1932,"children":1933},{"style":438},[1934],{"type":56,"value":669},{"type":50,"tag":416,"props":1936,"children":1937},{"style":423},[1938],{"type":56,"value":446},{"type":50,"tag":416,"props":1940,"children":1941},{"style":423},[1942],{"type":56,"value":451},{"type":50,"tag":416,"props":1944,"children":1945},{"style":423},[1946],{"type":56,"value":1947}," true,\n",{"type":50,"tag":416,"props":1949,"children":1950},{"class":418,"line":715},[1951,1955,1959,1963,1967,1971],{"type":50,"tag":416,"props":1952,"children":1953},{"style":423},[1954],{"type":56,"value":435},{"type":50,"tag":416,"props":1956,"children":1957},{"style":438},[1958],{"type":56,"value":694},{"type":50,"tag":416,"props":1960,"children":1961},{"style":423},[1962],{"type":56,"value":446},{"type":50,"tag":416,"props":1964,"children":1965},{"style":423},[1966],{"type":56,"value":451},{"type":50,"tag":416,"props":1968,"children":1969},{"style":705},[1970],{"type":56,"value":708},{"type":50,"tag":416,"props":1972,"children":1973},{"style":423},[1974],{"type":56,"value":471},{"type":50,"tag":416,"props":1976,"children":1977},{"class":418,"line":883},[1978,1982,1986,1990,1994],{"type":50,"tag":416,"props":1979,"children":1980},{"style":423},[1981],{"type":56,"value":435},{"type":50,"tag":416,"props":1983,"children":1984},{"style":438},[1985],{"type":56,"value":725},{"type":50,"tag":416,"props":1987,"children":1988},{"style":423},[1989],{"type":56,"value":446},{"type":50,"tag":416,"props":1991,"children":1992},{"style":423},[1993],{"type":56,"value":451},{"type":50,"tag":416,"props":1995,"children":1996},{"style":423},[1997],{"type":56,"value":1998}," [\n",{"type":50,"tag":416,"props":2000,"children":2001},{"class":418,"line":1134},[2002,2007,2011,2015,2019,2023,2027,2031,2035,2039,2043,2047,2051,2055,2059,2064,2068,2072,2076,2080,2084,2088,2092,2097,2101,2105,2109,2113,2117,2121,2125,2129,2133],{"type":50,"tag":416,"props":2003,"children":2004},{"style":423},[2005],{"type":56,"value":2006},"    {",{"type":50,"tag":416,"props":2008,"children":2009},{"style":423},[2010],{"type":56,"value":446},{"type":50,"tag":416,"props":2012,"children":2013},{"style":504},[2014],{"type":56,"value":747},{"type":50,"tag":416,"props":2016,"children":2017},{"style":423},[2018],{"type":56,"value":446},{"type":50,"tag":416,"props":2020,"children":2021},{"style":423},[2022],{"type":56,"value":451},{"type":50,"tag":416,"props":2024,"children":2025},{"style":423},[2026],{"type":56,"value":456},{"type":50,"tag":416,"props":2028,"children":2029},{"style":459},[2030],{"type":56,"value":764},{"type":50,"tag":416,"props":2032,"children":2033},{"style":423},[2034],{"type":56,"value":446},{"type":50,"tag":416,"props":2036,"children":2037},{"style":423},[2038],{"type":56,"value":623},{"type":50,"tag":416,"props":2040,"children":2041},{"style":423},[2042],{"type":56,"value":456},{"type":50,"tag":416,"props":2044,"children":2045},{"style":504},[2046],{"type":56,"value":781},{"type":50,"tag":416,"props":2048,"children":2049},{"style":423},[2050],{"type":56,"value":446},{"type":50,"tag":416,"props":2052,"children":2053},{"style":423},[2054],{"type":56,"value":451},{"type":50,"tag":416,"props":2056,"children":2057},{"style":423},[2058],{"type":56,"value":456},{"type":50,"tag":416,"props":2060,"children":2061},{"style":459},[2062],{"type":56,"value":2063},"$ai_model",{"type":50,"tag":416,"props":2065,"children":2066},{"style":423},[2067],{"type":56,"value":446},{"type":50,"tag":416,"props":2069,"children":2070},{"style":423},[2071],{"type":56,"value":623},{"type":50,"tag":416,"props":2073,"children":2074},{"style":423},[2075],{"type":56,"value":456},{"type":50,"tag":416,"props":2077,"children":2078},{"style":504},[2079],{"type":56,"value":814},{"type":50,"tag":416,"props":2081,"children":2082},{"style":423},[2083],{"type":56,"value":446},{"type":50,"tag":416,"props":2085,"children":2086},{"style":423},[2087],{"type":56,"value":451},{"type":50,"tag":416,"props":2089,"children":2090},{"style":423},[2091],{"type":56,"value":456},{"type":50,"tag":416,"props":2093,"children":2094},{"style":459},[2095],{"type":56,"value":2096},"gpt-4o",{"type":50,"tag":416,"props":2098,"children":2099},{"style":423},[2100],{"type":56,"value":446},{"type":50,"tag":416,"props":2102,"children":2103},{"style":423},[2104],{"type":56,"value":623},{"type":50,"tag":416,"props":2106,"children":2107},{"style":423},[2108],{"type":56,"value":456},{"type":50,"tag":416,"props":2110,"children":2111},{"style":504},[2112],{"type":56,"value":854},{"type":50,"tag":416,"props":2114,"children":2115},{"style":423},[2116],{"type":56,"value":446},{"type":50,"tag":416,"props":2118,"children":2119},{"style":423},[2120],{"type":56,"value":451},{"type":50,"tag":416,"props":2122,"children":2123},{"style":423},[2124],{"type":56,"value":456},{"type":50,"tag":416,"props":2126,"children":2127},{"style":459},[2128],{"type":56,"value":871},{"type":50,"tag":416,"props":2130,"children":2131},{"style":423},[2132],{"type":56,"value":446},{"type":50,"tag":416,"props":2134,"children":2135},{"style":423},[2136],{"type":56,"value":541},{"type":50,"tag":416,"props":2138,"children":2139},{"class":418,"line":1143},[2140],{"type":50,"tag":416,"props":2141,"children":2142},{"style":423},[2143],{"type":56,"value":2144},"  ]\n",{"type":50,"tag":416,"props":2146,"children":2147},{"class":418,"line":1185},[2148],{"type":50,"tag":416,"props":2149,"children":2150},{"style":423},[2151],{"type":56,"value":541},{"type":50,"tag":59,"props":2153,"children":2154},{},[2155],{"type":56,"value":2156},"Multiple filters are AND-ed together:",{"type":50,"tag":189,"props":2158,"children":2160},{"className":408,"code":2159,"language":410,"meta":194,"style":194},"posthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"$ai_provider\", \"value\": \"anthropic\", \"operator\": \"exact\"},\n    {\"type\": \"event\", \"key\": \"$ai_is_error\", \"value\": [\"true\"], \"operator\": \"exact\"}\n  ]\n}\n",[2161],{"type":50,"tag":106,"props":2162,"children":2163},{"__ignoreMap":194},[2164,2171,2178,2233,2256,2279,2416,2556,2563],{"type":50,"tag":416,"props":2165,"children":2166},{"class":418,"line":419},[2167],{"type":50,"tag":416,"props":2168,"children":2169},{"style":1147},[2170],{"type":56,"value":1860},{"type":50,"tag":416,"props":2172,"children":2173},{"class":418,"line":429},[2174],{"type":50,"tag":416,"props":2175,"children":2176},{"style":423},[2177],{"type":56,"value":426},{"type":50,"tag":416,"props":2179,"children":2180},{"class":418,"line":474},[2181,2185,2189,2193,2197,2201,2205,2209,2213,2217,2221,2225,2229],{"type":50,"tag":416,"props":2182,"children":2183},{"style":423},[2184],{"type":56,"value":435},{"type":50,"tag":416,"props":2186,"children":2187},{"style":438},[2188],{"type":56,"value":484},{"type":50,"tag":416,"props":2190,"children":2191},{"style":423},[2192],{"type":56,"value":446},{"type":50,"tag":416,"props":2194,"children":2195},{"style":423},[2196],{"type":56,"value":451},{"type":50,"tag":416,"props":2198,"children":2199},{"style":423},[2200],{"type":56,"value":497},{"type":50,"tag":416,"props":2202,"children":2203},{"style":423},[2204],{"type":56,"value":446},{"type":50,"tag":416,"props":2206,"children":2207},{"style":504},[2208],{"type":56,"value":327},{"type":50,"tag":416,"props":2210,"children":2211},{"style":423},[2212],{"type":56,"value":446},{"type":50,"tag":416,"props":2214,"children":2215},{"style":423},[2216],{"type":56,"value":451},{"type":50,"tag":416,"props":2218,"children":2219},{"style":423},[2220],{"type":56,"value":456},{"type":50,"tag":416,"props":2222,"children":2223},{"style":459},[2224],{"type":56,"value":919},{"type":50,"tag":416,"props":2226,"children":2227},{"style":423},[2228],{"type":56,"value":446},{"type":50,"tag":416,"props":2230,"children":2231},{"style":423},[2232],{"type":56,"value":1923},{"type":50,"tag":416,"props":2234,"children":2235},{"class":418,"line":535},[2236,2240,2244,2248,2252],{"type":50,"tag":416,"props":2237,"children":2238},{"style":423},[2239],{"type":56,"value":435},{"type":50,"tag":416,"props":2241,"children":2242},{"style":438},[2243],{"type":56,"value":669},{"type":50,"tag":416,"props":2245,"children":2246},{"style":423},[2247],{"type":56,"value":446},{"type":50,"tag":416,"props":2249,"children":2250},{"style":423},[2251],{"type":56,"value":451},{"type":50,"tag":416,"props":2253,"children":2254},{"style":423},[2255],{"type":56,"value":1947},{"type":50,"tag":416,"props":2257,"children":2258},{"class":418,"line":715},[2259,2263,2267,2271,2275],{"type":50,"tag":416,"props":2260,"children":2261},{"style":423},[2262],{"type":56,"value":435},{"type":50,"tag":416,"props":2264,"children":2265},{"style":438},[2266],{"type":56,"value":725},{"type":50,"tag":416,"props":2268,"children":2269},{"style":423},[2270],{"type":56,"value":446},{"type":50,"tag":416,"props":2272,"children":2273},{"style":423},[2274],{"type":56,"value":451},{"type":50,"tag":416,"props":2276,"children":2277},{"style":423},[2278],{"type":56,"value":1998},{"type":50,"tag":416,"props":2280,"children":2281},{"class":418,"line":883},[2282,2286,2290,2294,2298,2302,2306,2310,2314,2318,2322,2326,2330,2334,2338,2343,2347,2351,2355,2359,2363,2367,2371,2376,2380,2384,2388,2392,2396,2400,2404,2408,2412],{"type":50,"tag":416,"props":2283,"children":2284},{"style":423},[2285],{"type":56,"value":2006},{"type":50,"tag":416,"props":2287,"children":2288},{"style":423},[2289],{"type":56,"value":446},{"type":50,"tag":416,"props":2291,"children":2292},{"style":504},[2293],{"type":56,"value":747},{"type":50,"tag":416,"props":2295,"children":2296},{"style":423},[2297],{"type":56,"value":446},{"type":50,"tag":416,"props":2299,"children":2300},{"style":423},[2301],{"type":56,"value":451},{"type":50,"tag":416,"props":2303,"children":2304},{"style":423},[2305],{"type":56,"value":456},{"type":50,"tag":416,"props":2307,"children":2308},{"style":459},[2309],{"type":56,"value":764},{"type":50,"tag":416,"props":2311,"children":2312},{"style":423},[2313],{"type":56,"value":446},{"type":50,"tag":416,"props":2315,"children":2316},{"style":423},[2317],{"type":56,"value":623},{"type":50,"tag":416,"props":2319,"children":2320},{"style":423},[2321],{"type":56,"value":456},{"type":50,"tag":416,"props":2323,"children":2324},{"style":504},[2325],{"type":56,"value":781},{"type":50,"tag":416,"props":2327,"children":2328},{"style":423},[2329],{"type":56,"value":446},{"type":50,"tag":416,"props":2331,"children":2332},{"style":423},[2333],{"type":56,"value":451},{"type":50,"tag":416,"props":2335,"children":2336},{"style":423},[2337],{"type":56,"value":456},{"type":50,"tag":416,"props":2339,"children":2340},{"style":459},[2341],{"type":56,"value":2342},"$ai_provider",{"type":50,"tag":416,"props":2344,"children":2345},{"style":423},[2346],{"type":56,"value":446},{"type":50,"tag":416,"props":2348,"children":2349},{"style":423},[2350],{"type":56,"value":623},{"type":50,"tag":416,"props":2352,"children":2353},{"style":423},[2354],{"type":56,"value":456},{"type":50,"tag":416,"props":2356,"children":2357},{"style":504},[2358],{"type":56,"value":814},{"type":50,"tag":416,"props":2360,"children":2361},{"style":423},[2362],{"type":56,"value":446},{"type":50,"tag":416,"props":2364,"children":2365},{"style":423},[2366],{"type":56,"value":451},{"type":50,"tag":416,"props":2368,"children":2369},{"style":423},[2370],{"type":56,"value":456},{"type":50,"tag":416,"props":2372,"children":2373},{"style":459},[2374],{"type":56,"value":2375},"anthropic",{"type":50,"tag":416,"props":2377,"children":2378},{"style":423},[2379],{"type":56,"value":446},{"type":50,"tag":416,"props":2381,"children":2382},{"style":423},[2383],{"type":56,"value":623},{"type":50,"tag":416,"props":2385,"children":2386},{"style":423},[2387],{"type":56,"value":456},{"type":50,"tag":416,"props":2389,"children":2390},{"style":504},[2391],{"type":56,"value":854},{"type":50,"tag":416,"props":2393,"children":2394},{"style":423},[2395],{"type":56,"value":446},{"type":50,"tag":416,"props":2397,"children":2398},{"style":423},[2399],{"type":56,"value":451},{"type":50,"tag":416,"props":2401,"children":2402},{"style":423},[2403],{"type":56,"value":456},{"type":50,"tag":416,"props":2405,"children":2406},{"style":459},[2407],{"type":56,"value":871},{"type":50,"tag":416,"props":2409,"children":2410},{"style":423},[2411],{"type":56,"value":446},{"type":50,"tag":416,"props":2413,"children":2414},{"style":423},[2415],{"type":56,"value":1923},{"type":50,"tag":416,"props":2417,"children":2418},{"class":418,"line":1134},[2419,2423,2427,2431,2435,2439,2443,2447,2451,2455,2459,2463,2467,2471,2475,2479,2483,2487,2491,2495,2499,2503,2507,2511,2516,2520,2524,2528,2532,2536,2540,2544,2548,2552],{"type":50,"tag":416,"props":2420,"children":2421},{"style":423},[2422],{"type":56,"value":2006},{"type":50,"tag":416,"props":2424,"children":2425},{"style":423},[2426],{"type":56,"value":446},{"type":50,"tag":416,"props":2428,"children":2429},{"style":504},[2430],{"type":56,"value":747},{"type":50,"tag":416,"props":2432,"children":2433},{"style":423},[2434],{"type":56,"value":446},{"type":50,"tag":416,"props":2436,"children":2437},{"style":423},[2438],{"type":56,"value":451},{"type":50,"tag":416,"props":2440,"children":2441},{"style":423},[2442],{"type":56,"value":456},{"type":50,"tag":416,"props":2444,"children":2445},{"style":459},[2446],{"type":56,"value":764},{"type":50,"tag":416,"props":2448,"children":2449},{"style":423},[2450],{"type":56,"value":446},{"type":50,"tag":416,"props":2452,"children":2453},{"style":423},[2454],{"type":56,"value":623},{"type":50,"tag":416,"props":2456,"children":2457},{"style":423},[2458],{"type":56,"value":456},{"type":50,"tag":416,"props":2460,"children":2461},{"style":504},[2462],{"type":56,"value":781},{"type":50,"tag":416,"props":2464,"children":2465},{"style":423},[2466],{"type":56,"value":446},{"type":50,"tag":416,"props":2468,"children":2469},{"style":423},[2470],{"type":56,"value":451},{"type":50,"tag":416,"props":2472,"children":2473},{"style":423},[2474],{"type":56,"value":456},{"type":50,"tag":416,"props":2476,"children":2477},{"style":459},[2478],{"type":56,"value":1354},{"type":50,"tag":416,"props":2480,"children":2481},{"style":423},[2482],{"type":56,"value":446},{"type":50,"tag":416,"props":2484,"children":2485},{"style":423},[2486],{"type":56,"value":623},{"type":50,"tag":416,"props":2488,"children":2489},{"style":423},[2490],{"type":56,"value":456},{"type":50,"tag":416,"props":2492,"children":2493},{"style":504},[2494],{"type":56,"value":814},{"type":50,"tag":416,"props":2496,"children":2497},{"style":423},[2498],{"type":56,"value":446},{"type":50,"tag":416,"props":2500,"children":2501},{"style":423},[2502],{"type":56,"value":451},{"type":50,"tag":416,"props":2504,"children":2505},{"style":423},[2506],{"type":56,"value":827},{"type":50,"tag":416,"props":2508,"children":2509},{"style":423},[2510],{"type":56,"value":446},{"type":50,"tag":416,"props":2512,"children":2513},{"style":459},[2514],{"type":56,"value":2515},"true",{"type":50,"tag":416,"props":2517,"children":2518},{"style":423},[2519],{"type":56,"value":446},{"type":50,"tag":416,"props":2521,"children":2522},{"style":423},[2523],{"type":56,"value":845},{"type":50,"tag":416,"props":2525,"children":2526},{"style":423},[2527],{"type":56,"value":456},{"type":50,"tag":416,"props":2529,"children":2530},{"style":504},[2531],{"type":56,"value":854},{"type":50,"tag":416,"props":2533,"children":2534},{"style":423},[2535],{"type":56,"value":446},{"type":50,"tag":416,"props":2537,"children":2538},{"style":423},[2539],{"type":56,"value":451},{"type":50,"tag":416,"props":2541,"children":2542},{"style":423},[2543],{"type":56,"value":456},{"type":50,"tag":416,"props":2545,"children":2546},{"style":459},[2547],{"type":56,"value":871},{"type":50,"tag":416,"props":2549,"children":2550},{"style":423},[2551],{"type":56,"value":446},{"type":50,"tag":416,"props":2553,"children":2554},{"style":423},[2555],{"type":56,"value":541},{"type":50,"tag":416,"props":2557,"children":2558},{"class":418,"line":1143},[2559],{"type":50,"tag":416,"props":2560,"children":2561},{"style":423},[2562],{"type":56,"value":2144},{"type":50,"tag":416,"props":2564,"children":2565},{"class":418,"line":1185},[2566],{"type":50,"tag":416,"props":2567,"children":2568},{"style":423},[2569],{"type":56,"value":541},{"type":50,"tag":59,"props":2571,"children":2572},{},[2573,2575,2581,2582,2588,2589,2595],{"type":56,"value":2574},"You can also filter by person properties (discover them via ",{"type":50,"tag":106,"props":2576,"children":2578},{"className":2577},[],[2579],{"type":56,"value":2580},"read-data-schema",{"type":56,"value":1698},{"type":50,"tag":106,"props":2583,"children":2585},{"className":2584},[],[2586],{"type":56,"value":2587},"kind: \"entity_properties\"",{"type":56,"value":1448},{"type":50,"tag":106,"props":2590,"children":2592},{"className":2591},[],[2593],{"type":56,"value":2594},"entity: \"person\"",{"type":56,"value":2596},"):",{"type":50,"tag":189,"props":2598,"children":2600},{"className":408,"code":2599,"language":410,"meta":194,"style":194},"posthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-1h\"},\n  \"filterTestAccounts\": true,\n  \"properties\": [\n    {\"type\": \"person\", \"key\": \"email\", \"value\": \"@company.com\", \"operator\": \"icontains\"}\n  ]\n}\n",[2601],{"type":50,"tag":106,"props":2602,"children":2603},{"__ignoreMap":194},[2604,2611,2618,2673,2696,2719,2857,2864],{"type":50,"tag":416,"props":2605,"children":2606},{"class":418,"line":419},[2607],{"type":50,"tag":416,"props":2608,"children":2609},{"style":1147},[2610],{"type":56,"value":1860},{"type":50,"tag":416,"props":2612,"children":2613},{"class":418,"line":429},[2614],{"type":50,"tag":416,"props":2615,"children":2616},{"style":423},[2617],{"type":56,"value":426},{"type":50,"tag":416,"props":2619,"children":2620},{"class":418,"line":474},[2621,2625,2629,2633,2637,2641,2645,2649,2653,2657,2661,2665,2669],{"type":50,"tag":416,"props":2622,"children":2623},{"style":423},[2624],{"type":56,"value":435},{"type":50,"tag":416,"props":2626,"children":2627},{"style":438},[2628],{"type":56,"value":484},{"type":50,"tag":416,"props":2630,"children":2631},{"style":423},[2632],{"type":56,"value":446},{"type":50,"tag":416,"props":2634,"children":2635},{"style":423},[2636],{"type":56,"value":451},{"type":50,"tag":416,"props":2638,"children":2639},{"style":423},[2640],{"type":56,"value":497},{"type":50,"tag":416,"props":2642,"children":2643},{"style":423},[2644],{"type":56,"value":446},{"type":50,"tag":416,"props":2646,"children":2647},{"style":504},[2648],{"type":56,"value":327},{"type":50,"tag":416,"props":2650,"children":2651},{"style":423},[2652],{"type":56,"value":446},{"type":50,"tag":416,"props":2654,"children":2655},{"style":423},[2656],{"type":56,"value":451},{"type":50,"tag":416,"props":2658,"children":2659},{"style":423},[2660],{"type":56,"value":456},{"type":50,"tag":416,"props":2662,"children":2663},{"style":459},[2664],{"type":56,"value":919},{"type":50,"tag":416,"props":2666,"children":2667},{"style":423},[2668],{"type":56,"value":446},{"type":50,"tag":416,"props":2670,"children":2671},{"style":423},[2672],{"type":56,"value":1923},{"type":50,"tag":416,"props":2674,"children":2675},{"class":418,"line":535},[2676,2680,2684,2688,2692],{"type":50,"tag":416,"props":2677,"children":2678},{"style":423},[2679],{"type":56,"value":435},{"type":50,"tag":416,"props":2681,"children":2682},{"style":438},[2683],{"type":56,"value":669},{"type":50,"tag":416,"props":2685,"children":2686},{"style":423},[2687],{"type":56,"value":446},{"type":50,"tag":416,"props":2689,"children":2690},{"style":423},[2691],{"type":56,"value":451},{"type":50,"tag":416,"props":2693,"children":2694},{"style":423},[2695],{"type":56,"value":1947},{"type":50,"tag":416,"props":2697,"children":2698},{"class":418,"line":715},[2699,2703,2707,2711,2715],{"type":50,"tag":416,"props":2700,"children":2701},{"style":423},[2702],{"type":56,"value":435},{"type":50,"tag":416,"props":2704,"children":2705},{"style":438},[2706],{"type":56,"value":725},{"type":50,"tag":416,"props":2708,"children":2709},{"style":423},[2710],{"type":56,"value":446},{"type":50,"tag":416,"props":2712,"children":2713},{"style":423},[2714],{"type":56,"value":451},{"type":50,"tag":416,"props":2716,"children":2717},{"style":423},[2718],{"type":56,"value":1998},{"type":50,"tag":416,"props":2720,"children":2721},{"class":418,"line":883},[2722,2726,2730,2734,2738,2742,2746,2751,2755,2759,2763,2767,2771,2775,2779,2783,2787,2791,2795,2799,2803,2807,2811,2816,2820,2824,2828,2832,2836,2840,2844,2849,2853],{"type":50,"tag":416,"props":2723,"children":2724},{"style":423},[2725],{"type":56,"value":2006},{"type":50,"tag":416,"props":2727,"children":2728},{"style":423},[2729],{"type":56,"value":446},{"type":50,"tag":416,"props":2731,"children":2732},{"style":504},[2733],{"type":56,"value":747},{"type":50,"tag":416,"props":2735,"children":2736},{"style":423},[2737],{"type":56,"value":446},{"type":50,"tag":416,"props":2739,"children":2740},{"style":423},[2741],{"type":56,"value":451},{"type":50,"tag":416,"props":2743,"children":2744},{"style":423},[2745],{"type":56,"value":456},{"type":50,"tag":416,"props":2747,"children":2748},{"style":459},[2749],{"type":56,"value":2750},"person",{"type":50,"tag":416,"props":2752,"children":2753},{"style":423},[2754],{"type":56,"value":446},{"type":50,"tag":416,"props":2756,"children":2757},{"style":423},[2758],{"type":56,"value":623},{"type":50,"tag":416,"props":2760,"children":2761},{"style":423},[2762],{"type":56,"value":456},{"type":50,"tag":416,"props":2764,"children":2765},{"style":504},[2766],{"type":56,"value":781},{"type":50,"tag":416,"props":2768,"children":2769},{"style":423},[2770],{"type":56,"value":446},{"type":50,"tag":416,"props":2772,"children":2773},{"style":423},[2774],{"type":56,"value":451},{"type":50,"tag":416,"props":2776,"children":2777},{"style":423},[2778],{"type":56,"value":456},{"type":50,"tag":416,"props":2780,"children":2781},{"style":459},[2782],{"type":56,"value":1837},{"type":50,"tag":416,"props":2784,"children":2785},{"style":423},[2786],{"type":56,"value":446},{"type":50,"tag":416,"props":2788,"children":2789},{"style":423},[2790],{"type":56,"value":623},{"type":50,"tag":416,"props":2792,"children":2793},{"style":423},[2794],{"type":56,"value":456},{"type":50,"tag":416,"props":2796,"children":2797},{"style":504},[2798],{"type":56,"value":814},{"type":50,"tag":416,"props":2800,"children":2801},{"style":423},[2802],{"type":56,"value":446},{"type":50,"tag":416,"props":2804,"children":2805},{"style":423},[2806],{"type":56,"value":451},{"type":50,"tag":416,"props":2808,"children":2809},{"style":423},[2810],{"type":56,"value":456},{"type":50,"tag":416,"props":2812,"children":2813},{"style":459},[2814],{"type":56,"value":2815},"@company.com",{"type":50,"tag":416,"props":2817,"children":2818},{"style":423},[2819],{"type":56,"value":446},{"type":50,"tag":416,"props":2821,"children":2822},{"style":423},[2823],{"type":56,"value":623},{"type":50,"tag":416,"props":2825,"children":2826},{"style":423},[2827],{"type":56,"value":456},{"type":50,"tag":416,"props":2829,"children":2830},{"style":504},[2831],{"type":56,"value":854},{"type":50,"tag":416,"props":2833,"children":2834},{"style":423},[2835],{"type":56,"value":446},{"type":50,"tag":416,"props":2837,"children":2838},{"style":423},[2839],{"type":56,"value":451},{"type":50,"tag":416,"props":2841,"children":2842},{"style":423},[2843],{"type":56,"value":456},{"type":50,"tag":416,"props":2845,"children":2846},{"style":459},[2847],{"type":56,"value":2848},"icontains",{"type":50,"tag":416,"props":2850,"children":2851},{"style":423},[2852],{"type":56,"value":446},{"type":50,"tag":416,"props":2854,"children":2855},{"style":423},[2856],{"type":56,"value":541},{"type":50,"tag":416,"props":2858,"children":2859},{"class":418,"line":1134},[2860],{"type":50,"tag":416,"props":2861,"children":2862},{"style":423},[2863],{"type":56,"value":2144},{"type":50,"tag":416,"props":2865,"children":2866},{"class":418,"line":1143},[2867],{"type":50,"tag":416,"props":2868,"children":2869},{"style":423},[2870],{"type":56,"value":541},{"type":50,"tag":236,"props":2872,"children":2874},{"id":2873},"by-external-identifiers",[2875],{"type":56,"value":2876},"By external identifiers",{"type":50,"tag":59,"props":2878,"children":2879},{},[2880,2882,2887],{"type":56,"value":2881},"Customers often store their own IDs as event or person properties.\nUse ",{"type":50,"tag":106,"props":2883,"children":2885},{"className":2884},[],[2886],{"type":56,"value":145},{"type":56,"value":2888}," to discover what custom properties exist, then filter:",{"type":50,"tag":1298,"props":2890,"children":2891},{},[2892,2917,2922],{"type":50,"tag":252,"props":2893,"children":2894},{},[2895,2897,2902,2903,2908,2909,2915],{"type":56,"value":2896},"Call ",{"type":50,"tag":106,"props":2898,"children":2900},{"className":2899},[],[2901],{"type":56,"value":145},{"type":56,"value":1698},{"type":50,"tag":106,"props":2904,"children":2906},{"className":2905},[],[2907],{"type":56,"value":1734},{"type":56,"value":1448},{"type":50,"tag":106,"props":2910,"children":2912},{"className":2911},[],[2913],{"type":56,"value":2914},"event_name: \"$ai_trace\"",{"type":56,"value":2916}," to find custom properties",{"type":50,"tag":252,"props":2918,"children":2919},{},[2920],{"type":56,"value":2921},"Review the returned properties and their sample values",{"type":50,"tag":252,"props":2923,"children":2924},{},[2925],{"type":56,"value":2926},"Construct the filter using the discovered property key and a known value",{"type":50,"tag":189,"props":2928,"children":2930},{"className":408,"code":2929,"language":410,"meta":194,"style":194},"posthog:query-llm-traces-list\n{\n  \"dateRange\": {\"date_from\": \"-7d\"},\n  \"properties\": [\n    {\"type\": \"event\", \"key\": \"project_id\", \"value\": \"proj_abc123\", \"operator\": \"exact\"}\n  ]\n}\n",[2931],{"type":50,"tag":106,"props":2932,"children":2933},{"__ignoreMap":194},[2934,2941,2948,3003,3026,3162,3169],{"type":50,"tag":416,"props":2935,"children":2936},{"class":418,"line":419},[2937],{"type":50,"tag":416,"props":2938,"children":2939},{"style":1147},[2940],{"type":56,"value":1860},{"type":50,"tag":416,"props":2942,"children":2943},{"class":418,"line":429},[2944],{"type":50,"tag":416,"props":2945,"children":2946},{"style":423},[2947],{"type":56,"value":426},{"type":50,"tag":416,"props":2949,"children":2950},{"class":418,"line":474},[2951,2955,2959,2963,2967,2971,2975,2979,2983,2987,2991,2995,2999],{"type":50,"tag":416,"props":2952,"children":2953},{"style":423},[2954],{"type":56,"value":435},{"type":50,"tag":416,"props":2956,"children":2957},{"style":438},[2958],{"type":56,"value":484},{"type":50,"tag":416,"props":2960,"children":2961},{"style":423},[2962],{"type":56,"value":446},{"type":50,"tag":416,"props":2964,"children":2965},{"style":423},[2966],{"type":56,"value":451},{"type":50,"tag":416,"props":2968,"children":2969},{"style":423},[2970],{"type":56,"value":497},{"type":50,"tag":416,"props":2972,"children":2973},{"style":423},[2974],{"type":56,"value":446},{"type":50,"tag":416,"props":2976,"children":2977},{"style":504},[2978],{"type":56,"value":327},{"type":50,"tag":416,"props":2980,"children":2981},{"style":423},[2982],{"type":56,"value":446},{"type":50,"tag":416,"props":2984,"children":2985},{"style":423},[2986],{"type":56,"value":451},{"type":50,"tag":416,"props":2988,"children":2989},{"style":423},[2990],{"type":56,"value":456},{"type":50,"tag":416,"props":2992,"children":2993},{"style":459},[2994],{"type":56,"value":523},{"type":50,"tag":416,"props":2996,"children":2997},{"style":423},[2998],{"type":56,"value":446},{"type":50,"tag":416,"props":3000,"children":3001},{"style":423},[3002],{"type":56,"value":1923},{"type":50,"tag":416,"props":3004,"children":3005},{"class":418,"line":535},[3006,3010,3014,3018,3022],{"type":50,"tag":416,"props":3007,"children":3008},{"style":423},[3009],{"type":56,"value":435},{"type":50,"tag":416,"props":3011,"children":3012},{"style":438},[3013],{"type":56,"value":725},{"type":50,"tag":416,"props":3015,"children":3016},{"style":423},[3017],{"type":56,"value":446},{"type":50,"tag":416,"props":3019,"children":3020},{"style":423},[3021],{"type":56,"value":451},{"type":50,"tag":416,"props":3023,"children":3024},{"style":423},[3025],{"type":56,"value":1998},{"type":50,"tag":416,"props":3027,"children":3028},{"class":418,"line":715},[3029,3033,3037,3041,3045,3049,3053,3057,3061,3065,3069,3073,3077,3081,3085,3089,3093,3097,3101,3105,3109,3113,3117,3122,3126,3130,3134,3138,3142,3146,3150,3154,3158],{"type":50,"tag":416,"props":3030,"children":3031},{"style":423},[3032],{"type":56,"value":2006},{"type":50,"tag":416,"props":3034,"children":3035},{"style":423},[3036],{"type":56,"value":446},{"type":50,"tag":416,"props":3038,"children":3039},{"style":504},[3040],{"type":56,"value":747},{"type":50,"tag":416,"props":3042,"children":3043},{"style":423},[3044],{"type":56,"value":446},{"type":50,"tag":416,"props":3046,"children":3047},{"style":423},[3048],{"type":56,"value":451},{"type":50,"tag":416,"props":3050,"children":3051},{"style":423},[3052],{"type":56,"value":456},{"type":50,"tag":416,"props":3054,"children":3055},{"style":459},[3056],{"type":56,"value":764},{"type":50,"tag":416,"props":3058,"children":3059},{"style":423},[3060],{"type":56,"value":446},{"type":50,"tag":416,"props":3062,"children":3063},{"style":423},[3064],{"type":56,"value":623},{"type":50,"tag":416,"props":3066,"children":3067},{"style":423},[3068],{"type":56,"value":456},{"type":50,"tag":416,"props":3070,"children":3071},{"style":504},[3072],{"type":56,"value":781},{"type":50,"tag":416,"props":3074,"children":3075},{"style":423},[3076],{"type":56,"value":446},{"type":50,"tag":416,"props":3078,"children":3079},{"style":423},[3080],{"type":56,"value":451},{"type":50,"tag":416,"props":3082,"children":3083},{"style":423},[3084],{"type":56,"value":456},{"type":50,"tag":416,"props":3086,"children":3087},{"style":459},[3088],{"type":56,"value":1803},{"type":50,"tag":416,"props":3090,"children":3091},{"style":423},[3092],{"type":56,"value":446},{"type":50,"tag":416,"props":3094,"children":3095},{"style":423},[3096],{"type":56,"value":623},{"type":50,"tag":416,"props":3098,"children":3099},{"style":423},[3100],{"type":56,"value":456},{"type":50,"tag":416,"props":3102,"children":3103},{"style":504},[3104],{"type":56,"value":814},{"type":50,"tag":416,"props":3106,"children":3107},{"style":423},[3108],{"type":56,"value":446},{"type":50,"tag":416,"props":3110,"children":3111},{"style":423},[3112],{"type":56,"value":451},{"type":50,"tag":416,"props":3114,"children":3115},{"style":423},[3116],{"type":56,"value":456},{"type":50,"tag":416,"props":3118,"children":3119},{"style":459},[3120],{"type":56,"value":3121},"proj_abc123",{"type":50,"tag":416,"props":3123,"children":3124},{"style":423},[3125],{"type":56,"value":446},{"type":50,"tag":416,"props":3127,"children":3128},{"style":423},[3129],{"type":56,"value":623},{"type":50,"tag":416,"props":3131,"children":3132},{"style":423},[3133],{"type":56,"value":456},{"type":50,"tag":416,"props":3135,"children":3136},{"style":504},[3137],{"type":56,"value":854},{"type":50,"tag":416,"props":3139,"children":3140},{"style":423},[3141],{"type":56,"value":446},{"type":50,"tag":416,"props":3143,"children":3144},{"style":423},[3145],{"type":56,"value":451},{"type":50,"tag":416,"props":3147,"children":3148},{"style":423},[3149],{"type":56,"value":456},{"type":50,"tag":416,"props":3151,"children":3152},{"style":459},[3153],{"type":56,"value":871},{"type":50,"tag":416,"props":3155,"children":3156},{"style":423},[3157],{"type":56,"value":446},{"type":50,"tag":416,"props":3159,"children":3160},{"style":423},[3161],{"type":56,"value":541},{"type":50,"tag":416,"props":3163,"children":3164},{"class":418,"line":883},[3165],{"type":50,"tag":416,"props":3166,"children":3167},{"style":423},[3168],{"type":56,"value":2144},{"type":50,"tag":416,"props":3170,"children":3171},{"class":418,"line":1134},[3172],{"type":50,"tag":416,"props":3173,"children":3174},{"style":423},[3175],{"type":56,"value":541},{"type":50,"tag":59,"props":3177,"children":3178},{},[3179],{"type":56,"value":3180},"For more complex SQL patterns, read these references:",{"type":50,"tag":248,"props":3182,"children":3183},{},[3184,3203],{"type":50,"tag":252,"props":3185,"children":3186},{},[3187,3193,3195,3201],{"type":50,"tag":180,"props":3188,"children":3190},{"href":3189},".\u002Freferences\u002Fexample-llm-trace.md.j2",[3191],{"type":56,"value":3192},"Single trace retrieval",{"type":56,"value":3194}," — fetches a single trace by ID with all events and properties (renders the ",{"type":50,"tag":106,"props":3196,"children":3198},{"className":3197},[],[3199],{"type":56,"value":3200},"TraceQuery",{"type":56,"value":3202}," HogQL)",{"type":50,"tag":252,"props":3204,"children":3205},{},[3206,3212],{"type":50,"tag":180,"props":3207,"children":3209},{"href":3208},".\u002Freferences\u002Fexample-llm-traces-list.md",[3210],{"type":56,"value":3211},"Traces list with aggregated metrics",{"type":56,"value":3213}," — two-phase query: find trace IDs first, then fetch aggregated latency, tokens, costs, and error counts",{"type":50,"tag":65,"props":3215,"children":3217},{"id":3216},"parsing-large-trace-results",[3218],{"type":56,"value":3219},"Parsing large trace results",{"type":50,"tag":59,"props":3221,"children":3222},{},[3223],{"type":56,"value":3224},"Trace tool results are JSON. When too large to read inline, Claude Code persists them to a file.",{"type":50,"tag":236,"props":3226,"children":3228},{"id":3227},"persisted-file-format",[3229],{"type":56,"value":3230},"Persisted file format",{"type":50,"tag":189,"props":3232,"children":3234},{"className":408,"code":3233,"language":410,"meta":194,"style":194},"[{ \"type\": \"text\", \"text\": \"{\\\"results\\\": [...], \\\"_posthogUrl\\\": \\\"...\\\"}\" }]\n",[3235],{"type":50,"tag":106,"props":3236,"children":3237},{"__ignoreMap":194},[3238],{"type":50,"tag":416,"props":3239,"children":3240},{"class":418,"line":419},[3241,3246,3250,3254,3258,3262,3266,3270,3274,3278,3282,3286,3290,3294,3298,3303,3308,3313,3317,3322,3326,3330,3334,3338,3342,3347,3351,3356,3360],{"type":50,"tag":416,"props":3242,"children":3243},{"style":423},[3244],{"type":56,"value":3245},"[{",{"type":50,"tag":416,"props":3247,"children":3248},{"style":423},[3249],{"type":56,"value":456},{"type":50,"tag":416,"props":3251,"children":3252},{"style":438},[3253],{"type":56,"value":747},{"type":50,"tag":416,"props":3255,"children":3256},{"style":423},[3257],{"type":56,"value":446},{"type":50,"tag":416,"props":3259,"children":3260},{"style":423},[3261],{"type":56,"value":451},{"type":50,"tag":416,"props":3263,"children":3264},{"style":423},[3265],{"type":56,"value":456},{"type":50,"tag":416,"props":3267,"children":3268},{"style":459},[3269],{"type":56,"value":56},{"type":50,"tag":416,"props":3271,"children":3272},{"style":423},[3273],{"type":56,"value":446},{"type":50,"tag":416,"props":3275,"children":3276},{"style":423},[3277],{"type":56,"value":623},{"type":50,"tag":416,"props":3279,"children":3280},{"style":423},[3281],{"type":56,"value":456},{"type":50,"tag":416,"props":3283,"children":3284},{"style":438},[3285],{"type":56,"value":56},{"type":50,"tag":416,"props":3287,"children":3288},{"style":423},[3289],{"type":56,"value":446},{"type":50,"tag":416,"props":3291,"children":3292},{"style":423},[3293],{"type":56,"value":451},{"type":50,"tag":416,"props":3295,"children":3296},{"style":423},[3297],{"type":56,"value":456},{"type":50,"tag":416,"props":3299,"children":3300},{"style":459},[3301],{"type":56,"value":3302},"{",{"type":50,"tag":416,"props":3304,"children":3305},{"style":1147},[3306],{"type":56,"value":3307},"\\\"",{"type":50,"tag":416,"props":3309,"children":3310},{"style":459},[3311],{"type":56,"value":3312},"results",{"type":50,"tag":416,"props":3314,"children":3315},{"style":1147},[3316],{"type":56,"value":3307},{"type":50,"tag":416,"props":3318,"children":3319},{"style":459},[3320],{"type":56,"value":3321},": [...], ",{"type":50,"tag":416,"props":3323,"children":3324},{"style":1147},[3325],{"type":56,"value":3307},{"type":50,"tag":416,"props":3327,"children":3328},{"style":459},[3329],{"type":56,"value":1000},{"type":50,"tag":416,"props":3331,"children":3332},{"style":1147},[3333],{"type":56,"value":3307},{"type":50,"tag":416,"props":3335,"children":3336},{"style":459},[3337],{"type":56,"value":1546},{"type":50,"tag":416,"props":3339,"children":3340},{"style":1147},[3341],{"type":56,"value":3307},{"type":50,"tag":416,"props":3343,"children":3344},{"style":459},[3345],{"type":56,"value":3346},"...",{"type":50,"tag":416,"props":3348,"children":3349},{"style":1147},[3350],{"type":56,"value":3307},{"type":50,"tag":416,"props":3352,"children":3353},{"style":459},[3354],{"type":56,"value":3355},"}",{"type":50,"tag":416,"props":3357,"children":3358},{"style":423},[3359],{"type":56,"value":446},{"type":50,"tag":416,"props":3361,"children":3362},{"style":423},[3363],{"type":56,"value":880},{"type":50,"tag":236,"props":3365,"children":3367},{"id":3366},"trace-json-structure",[3368],{"type":56,"value":3369},"Trace JSON structure",{"type":50,"tag":189,"props":3371,"children":3374},{"className":3372,"code":3373,"language":56,"meta":194},[192],"results (array for list, object for single trace)\n  ├── id, traceName, createdAt, totalLatency, totalCost\n  ├── inputState, outputState (trace-level state)\n  └── events[]\n        ├── event ($ai_span | $ai_generation | $ai_embedding | $ai_metric | $ai_feedback)\n        ├── id, createdAt\n        └── properties\n              ├── $ai_span_name, $ai_latency, $ai_is_error\n              ├── $ai_input_state, $ai_output_state (span tool I\u002FO)\n              ├── $ai_input, $ai_output_choices (generation messages)\n              ├── $ai_model, $ai_provider\n              └── $ai_input_tokens, $ai_output_tokens, $ai_total_cost_usd\n",[3375],{"type":50,"tag":106,"props":3376,"children":3377},{"__ignoreMap":194},[3378],{"type":56,"value":3373},{"type":50,"tag":236,"props":3380,"children":3382},{"id":3381},"available-scripts",[3383],{"type":56,"value":3384},"Available scripts",{"type":50,"tag":72,"props":3386,"children":3387},{},[3388,3408],{"type":50,"tag":76,"props":3389,"children":3390},{},[3391],{"type":50,"tag":80,"props":3392,"children":3393},{},[3394,3399,3403],{"type":50,"tag":84,"props":3395,"children":3396},{},[3397],{"type":56,"value":3398},"Script",{"type":50,"tag":84,"props":3400,"children":3401},{},[3402],{"type":56,"value":93},{"type":50,"tag":84,"props":3404,"children":3405},{},[3406],{"type":56,"value":3407},"Usage",{"type":50,"tag":95,"props":3409,"children":3410},{},[3411,3441,3471,3501,3531,3560],{"type":50,"tag":80,"props":3412,"children":3413},{},[3414,3427,3432],{"type":50,"tag":102,"props":3415,"children":3416},{},[3417],{"type":50,"tag":180,"props":3418,"children":3420},{"href":3419},".\u002Fscripts\u002Fprint_summary.py",[3421],{"type":50,"tag":106,"props":3422,"children":3424},{"className":3423},[],[3425],{"type":56,"value":3426},"print_summary.py",{"type":50,"tag":102,"props":3428,"children":3429},{},[3430],{"type":56,"value":3431},"Aggregate list\u002Fsession totals, trace metadata, tool calls, errors, and final LLM output",{"type":50,"tag":102,"props":3433,"children":3434},{},[3435],{"type":50,"tag":106,"props":3436,"children":3438},{"className":3437},[],[3439],{"type":56,"value":3440},"python3 scripts\u002Fprint_summary.py FILE",{"type":50,"tag":80,"props":3442,"children":3443},{},[3444,3457,3462],{"type":50,"tag":102,"props":3445,"children":3446},{},[3447],{"type":50,"tag":180,"props":3448,"children":3450},{"href":3449},".\u002Fscripts\u002Fprint_timeline.py",[3451],{"type":50,"tag":106,"props":3452,"children":3454},{"className":3453},[],[3455],{"type":56,"value":3456},"print_timeline.py",{"type":50,"tag":102,"props":3458,"children":3459},{},[3460],{"type":56,"value":3461},"Chronological event timeline with I\u002FO summaries",{"type":50,"tag":102,"props":3463,"children":3464},{},[3465],{"type":50,"tag":106,"props":3466,"children":3468},{"className":3467},[],[3469],{"type":56,"value":3470},"python3 scripts\u002Fprint_timeline.py FILE",{"type":50,"tag":80,"props":3472,"children":3473},{},[3474,3487,3492],{"type":50,"tag":102,"props":3475,"children":3476},{},[3477],{"type":50,"tag":180,"props":3478,"children":3480},{"href":3479},".\u002Fscripts\u002Fextract_span.py",[3481],{"type":50,"tag":106,"props":3482,"children":3484},{"className":3483},[],[3485],{"type":56,"value":3486},"extract_span.py",{"type":50,"tag":102,"props":3488,"children":3489},{},[3490],{"type":56,"value":3491},"Full input\u002Foutput of a specific span by name",{"type":50,"tag":102,"props":3493,"children":3494},{},[3495],{"type":50,"tag":106,"props":3496,"children":3498},{"className":3497},[],[3499],{"type":56,"value":3500},"SPAN=\"name\" python3 scripts\u002Fextract_span.py FILE",{"type":50,"tag":80,"props":3502,"children":3503},{},[3504,3517,3522],{"type":50,"tag":102,"props":3505,"children":3506},{},[3507],{"type":50,"tag":180,"props":3508,"children":3510},{"href":3509},".\u002Fscripts\u002Fextract_conversation.py",[3511],{"type":50,"tag":106,"props":3512,"children":3514},{"className":3513},[],[3515],{"type":56,"value":3516},"extract_conversation.py",{"type":50,"tag":102,"props":3518,"children":3519},{},[3520],{"type":56,"value":3521},"LLM messages with thinking blocks and tool calls",{"type":50,"tag":102,"props":3523,"children":3524},{},[3525],{"type":50,"tag":106,"props":3526,"children":3528},{"className":3527},[],[3529],{"type":56,"value":3530},"python3 scripts\u002Fextract_conversation.py FILE",{"type":50,"tag":80,"props":3532,"children":3533},{},[3534,3546,3551],{"type":50,"tag":102,"props":3535,"children":3536},{},[3537],{"type":50,"tag":180,"props":3538,"children":3540},{"href":3539},".\u002Fscripts\u002Fsearch_traces.py",[3541],{"type":50,"tag":106,"props":3542,"children":3544},{"className":3543},[],[3545],{"type":56,"value":1485},{"type":50,"tag":102,"props":3547,"children":3548},{},[3549],{"type":56,"value":3550},"Find a keyword across all event properties",{"type":50,"tag":102,"props":3552,"children":3553},{},[3554],{"type":50,"tag":106,"props":3555,"children":3557},{"className":3556},[],[3558],{"type":56,"value":3559},"SEARCH=\"keyword\" python3 scripts\u002Fsearch_traces.py FILE",{"type":50,"tag":80,"props":3561,"children":3562},{},[3563,3576,3581],{"type":50,"tag":102,"props":3564,"children":3565},{},[3566],{"type":50,"tag":180,"props":3567,"children":3569},{"href":3568},".\u002Fscripts\u002Fshow_structure.py",[3570],{"type":50,"tag":106,"props":3571,"children":3573},{"className":3572},[],[3574],{"type":56,"value":3575},"show_structure.py",{"type":50,"tag":102,"props":3577,"children":3578},{},[3579],{"type":56,"value":3580},"Show JSON keys and types without values",{"type":50,"tag":102,"props":3582,"children":3583},{},[3584],{"type":50,"tag":106,"props":3585,"children":3587},{"className":3586},[],[3588],{"type":56,"value":3589},"cat blob.json | python3 scripts\u002Fshow_structure.py",{"type":50,"tag":65,"props":3591,"children":3593},{"id":3592},"tips",[3594],{"type":56,"value":3595},"Tips",{"type":50,"tag":248,"props":3597,"children":3598},{},[3599,3639,3651,3667,3683,3749,3761],{"type":50,"tag":252,"props":3600,"children":3601},{},[3602,3604,3609,3611,3617,3618,3623,3625,3630,3631,3637],{"type":56,"value":3603},"Always set ",{"type":50,"tag":106,"props":3605,"children":3607},{"className":3606},[],[3608],{"type":56,"value":484},{"type":56,"value":3610}," — queries without a time range are slow. Use narrow windows (",{"type":50,"tag":106,"props":3612,"children":3614},{"className":3613},[],[3615],{"type":56,"value":3616},"-30m",{"type":56,"value":955},{"type":50,"tag":106,"props":3619,"children":3621},{"className":3620},[],[3622],{"type":56,"value":919},{"type":56,"value":3624},") for broad listing queries; wider windows (",{"type":50,"tag":106,"props":3626,"children":3628},{"className":3627},[],[3629],{"type":56,"value":523},{"type":56,"value":955},{"type":50,"tag":106,"props":3632,"children":3634},{"className":3633},[],[3635],{"type":56,"value":3636},"-30d",{"type":56,"value":3638},") are fine for narrow queries filtered by trace ID or specific property values",{"type":50,"tag":252,"props":3640,"children":3641},{},[3642,3644,3649],{"type":56,"value":3643},"Always include the ",{"type":50,"tag":106,"props":3645,"children":3647},{"className":3646},[],[3648],{"type":56,"value":1000},{"type":56,"value":3650}," in your response so the user can click through",{"type":50,"tag":252,"props":3652,"children":3653},{},[3654,3659,3660,3665],{"type":50,"tag":106,"props":3655,"children":3657},{"className":3656},[],[3658],{"type":56,"value":1330},{"type":56,"value":270},{"type":50,"tag":106,"props":3661,"children":3663},{"className":3662},[],[3664],{"type":56,"value":1342},{"type":56,"value":3666}," on spans contain tool call inputs and outputs",{"type":50,"tag":252,"props":3668,"children":3669},{},[3670,3675,3676,3681],{"type":50,"tag":106,"props":3671,"children":3673},{"className":3672},[],[3674],{"type":56,"value":1027},{"type":56,"value":270},{"type":50,"tag":106,"props":3677,"children":3679},{"className":3678},[],[3680],{"type":56,"value":1035},{"type":56,"value":3682}," on generations contain the full LLM conversation — can be megabytes; when the result is persisted to a file, use the parsing scripts",{"type":50,"tag":252,"props":3684,"children":3685},{},[3686,3688,3693,3694,3700,3701,3706,3707,3712,3713,3718,3719,3725,3727,3733,3735,3741,3743,3747],{"type":56,"value":3687},"In raw SQL, heavy content (",{"type":50,"tag":106,"props":3689,"children":3691},{"className":3690},[],[3692],{"type":56,"value":1027},{"type":56,"value":270},{"type":50,"tag":106,"props":3695,"children":3697},{"className":3696},[],[3698],{"type":56,"value":3699},"$ai_output",{"type":56,"value":270},{"type":50,"tag":106,"props":3702,"children":3704},{"className":3703},[],[3705],{"type":56,"value":1035},{"type":56,"value":270},{"type":50,"tag":106,"props":3708,"children":3710},{"className":3709},[],[3711],{"type":56,"value":1330},{"type":56,"value":270},{"type":50,"tag":106,"props":3714,"children":3716},{"className":3715},[],[3717],{"type":56,"value":1342},{"type":56,"value":270},{"type":50,"tag":106,"props":3720,"children":3722},{"className":3721},[],[3723],{"type":56,"value":3724},"$ai_tools",{"type":56,"value":3726},") lives only on the ",{"type":50,"tag":106,"props":3728,"children":3730},{"className":3729},[],[3731],{"type":56,"value":3732},"posthog.ai_events",{"type":56,"value":3734}," table, not ",{"type":50,"tag":106,"props":3736,"children":3738},{"className":3737},[],[3739],{"type":56,"value":3740},"events.properties",{"type":56,"value":3742}," — see the ",{"type":50,"tag":180,"props":3744,"children":3745},{"href":182},[3746],{"type":56,"value":185},{"type":56,"value":3748}," for the column mapping and trace-id-anchored query patterns",{"type":50,"tag":252,"props":3750,"children":3751},{},[3752,3753,3759],{"type":56,"value":1479},{"type":50,"tag":106,"props":3754,"children":3756},{"className":3755},[],[3757],{"type":56,"value":3758},"filterTestAccounts: true",{"type":56,"value":3760}," to exclude internal\u002Ftest traffic when searching",{"type":50,"tag":252,"props":3762,"children":3763},{},[3764,3770,3772,3778,3780,3786,3787,3793,3794],{"type":50,"tag":106,"props":3765,"children":3767},{"className":3766},[],[3768],{"type":56,"value":3769},"$ai_trace",{"type":56,"value":3771}," events are NOT in the ",{"type":50,"tag":106,"props":3773,"children":3775},{"className":3774},[],[3776],{"type":56,"value":3777},"events",{"type":56,"value":3779}," array — their data is surfaced via trace-level ",{"type":50,"tag":106,"props":3781,"children":3783},{"className":3782},[],[3784],{"type":56,"value":3785},"inputState",{"type":56,"value":955},{"type":50,"tag":106,"props":3788,"children":3790},{"className":3789},[],[3791],{"type":56,"value":3792},"outputState",{"type":56,"value":1772},{"type":50,"tag":106,"props":3795,"children":3797},{"className":3796},[],[3798],{"type":56,"value":3799},"traceName",{"type":50,"tag":3801,"props":3802,"children":3803},"style",{},[3804],{"type":56,"value":3805},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":3807,"total":3923},[3808,3825,3843,3860,3873,3887,3905],{"slug":3809,"name":3809,"fn":3810,"description":3811,"org":3812,"tags":3813,"stars":29,"repoUrl":30,"updatedAt":3824},"analyzing-experiment-session-replays","analyze session replays for PostHog experiments","Analyze session replay patterns across experiment variants to understand user behavior differences. Use when the user wants to see how users interact with different experiment variants, identify usability issues, compare behavior patterns between control and test groups, or get qualitative insights to complement quantitative experiment results.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3814,3817,3820,3821],{"name":3815,"slug":3816,"type":15},"Analytics","analytics",{"name":3818,"slug":3819,"type":15},"Design","design",{"name":9,"slug":8,"type":15},{"name":3822,"slug":3823,"type":15},"User Research","user-research","2026-04-06T18:44:38.291781",{"slug":3826,"name":3826,"fn":3827,"description":3828,"org":3829,"tags":3830,"stars":29,"repoUrl":30,"updatedAt":3842},"assessing-heatmaps","analyze page heatmaps and suggest improvements","Assesses what a page's heatmap is telling you and recommends concrete changes. Pulls click \u002F rageclick \u002F scroll-depth data for a URL, names the hot elements by cross-referencing autocapture events on the same page, and can create a saved heatmap the user opens in PostHog, then summarizes the behavior and proposes improvements.\nTRIGGER when: user asks what a heatmap shows, why people aren't clicking something, where users rage-click, how far they scroll, what to change on a page based on heatmap\u002Fclick data, or to 'analyze\u002Fassess\u002Freview the heatmap' for a URL.\nDO NOT TRIGGER when: the user only wants to create a saved heatmap screenshot with no analysis (use heatmaps-saved-create directly), or is asking about session replay in general (use investigating-replay).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3831,3832,3835,3836,3839],{"name":3815,"slug":3816,"type":15},{"name":3833,"slug":3834,"type":15},"Frontend","frontend",{"name":9,"slug":8,"type":15},{"name":3837,"slug":3838,"type":15},"Product Management","product-management",{"name":3840,"slug":3841,"type":15},"UX Design","ux-design","2026-06-05T07:40:43.37798",{"slug":3844,"name":3844,"fn":3845,"description":3846,"org":3847,"tags":3848,"stars":29,"repoUrl":30,"updatedAt":3859},"auditing-experiments-flags","audit PostHog experiments and feature flags","Audit PostHog experiments and feature flags for configuration issues, staleness, and best-practice violations. Read when the user asks to audit, health-check, or review experiments or feature flags, check flag hygiene, or verify experiment setup.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3849,3852,3855,3856],{"name":3850,"slug":3851,"type":15},"Audit","audit",{"name":3853,"slug":3854,"type":15},"Feature Flags","feature-flags",{"name":9,"slug":8,"type":15},{"name":3857,"slug":3858,"type":15},"QA","qa","2026-04-06T18:44:30.657553",{"slug":3861,"name":3861,"fn":3862,"description":3863,"org":3864,"tags":3865,"stars":29,"repoUrl":30,"updatedAt":3872},"authoring-scouts","author and edit PostHog Signals scouts","How to author, edit, and adapt PostHog Signals scouts — the scheduled agents that scan a project and write reports into the Signals inbox. Use when a user wants to customize a canonical scout for their own setup (narrow its scope, retune its thresholds, add disqualifiers), tweak a scout's schedule or dry-run posture, or write a brand-new scout from scratch for a specific use case (a custom event, a product surface no canonical scout covers), or steer a scout without editing it at all by leaving it a note. Covers the scout SKILL.md anatomy, the report contract, the dedupe + scratchpad-memory conventions, the scout-notes steering channel, the per-team skills-store path vs the canonical in-repo path, and the write-and-inspect test loop (with dry-run as an optional safety net). Trigger on \"write\u002Fedit\u002Fcustomize a signals scout\", \"new scout for X\", \"tune my scout schedule\", \"make a scout that watches \u003Cevent>\", \"leave a note for \u002F give feedback to a scout\", \"tell the scouts about X\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3866,3867,3870,3871],{"name":24,"slug":25,"type":15},{"name":3868,"slug":3869,"type":15},"Automation","automation",{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},"2026-07-28T05:33:45.509154",{"slug":3874,"name":3874,"fn":3875,"description":3876,"org":3877,"tags":3878,"stars":29,"repoUrl":30,"updatedAt":3886},"building-a-dashboard","build and update PostHog dashboards","Build a new dashboard, or update an existing one, from a set of insights — the same job the in-app assistant does with its upsert-dashboard tool, but over MCP. Use when a user asks to create a dashboard, put several metrics\u002Fcharts together on one page, assemble a dashboard for a topic (product analytics, retention, revenue, activation, etc.), or add\u002Fremove\u002Freplace insights on a dashboard they already have. Covers deciding create vs update, reusing existing insights vs creating new ones, and using PostHog's vetted dashboard templates as reference for what a strong dashboard on a topic looks like.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3879,3880,3883],{"name":3815,"slug":3816,"type":15},{"name":3881,"slug":3882,"type":15},"Dashboards","dashboards",{"name":3884,"slug":3885,"type":15},"MCP","mcp","2026-07-21T06:07:38.060598",{"slug":3888,"name":3888,"fn":3889,"description":3890,"org":3891,"tags":3892,"stars":29,"repoUrl":30,"updatedAt":3904},"checking-deploy-timing","correlate PostHog deployments with GitHub commits","Determine when a PostHog code change reached a given environment by reading the hidden GIT deploy annotations in the project and correlating them with the merge commit on GitHub. Use when PostHog staff ask \"when was X deployed\", \"is my change live in the US\u002FEU yet\", \"has my PR shipped\", \"did the fix roll out to prod-us\", or otherwise want to know whether\u002Fwhen a commit, PR, or feature went out to a region. Do not answer deploy-timing questions from event\u002Fdata volume alone — that only shows when data changed, not when code shipped.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3893,3896,3899,3902,3903],{"name":3894,"slug":3895,"type":15},"Deployment","deployment",{"name":3897,"slug":3898,"type":15},"Git","git",{"name":3900,"slug":3901,"type":15},"GitHub","github",{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},"2026-06-28T07:46:59.53536",{"slug":3906,"name":3906,"fn":3907,"description":3908,"org":3909,"tags":3910,"stars":29,"repoUrl":30,"updatedAt":3922},"choosing-trend-or-slope-view","visualize trends and growth over time","Clarify how to visualize change over a time range before building a trend. Use whenever the user asks how much something changed, grew, dropped, improved, or regressed between two points or periods — \"how much did X change from A to B\", \"before vs after\", \"start vs end\", \"week over week\", \"compare this month to last\", \"change over time\" — or mentions a \"slope chart\" \u002F \"slopegraph\". Two readings of \"change\" need different charts: the whole trend (a line, every interval) versus just the two endpoints (a slope, start vs end). Ask which they want, then render it. Not for choosing a saved insight ChartDisplayType in the insight editor.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3911,3912,3915,3918,3919],{"name":3815,"slug":3816,"type":15},{"name":3913,"slug":3914,"type":15},"Charts","charts",{"name":3916,"slug":3917,"type":15},"Data Visualization","data-visualization",{"name":9,"slug":8,"type":15},{"name":3920,"slug":3921,"type":15},"Reporting","reporting","2026-06-18T08:18:57.960157",56,{"items":3925,"total":4081},[3926,3941,3951,3964,3977,3990,4006,4019,4031,4046,4056,4071],{"slug":3927,"name":3927,"fn":3928,"description":3929,"org":3930,"tags":3931,"stars":3938,"repoUrl":3939,"updatedAt":3940},"analyzing-expensive-users","analyze expensive users in AI observability","Analyze the most expensive users in AI observability and explain why they cost so much. Use when the user asks about top spenders, expensive users, per-user LLM cost, user-level cost drivers, or patterns behind high AI observability spend.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3932,3933,3936,3937],{"name":3815,"slug":3816,"type":15},{"name":3934,"slug":3935,"type":15},"Cost Optimization","cost-optimization",{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},35568,"https:\u002F\u002Fgithub.com\u002FPostHog\u002Fposthog","2026-07-28T05:34:11.117757",{"slug":3942,"name":3942,"fn":3943,"description":3944,"org":3945,"tags":3946,"stars":3938,"repoUrl":3939,"updatedAt":3950},"auditing-endpoints","audit PostHog project endpoints","Audit every endpoint in a PostHog project for staleness, failed materialisations, and unused materialised versions. Use when the user asks \"what endpoints can I clean up?\", \"are any of my endpoints broken?\", \"which materialised versions are still being called?\", or wants a one-shot cleanup pass over the Endpoints product. Produces a prioritised report grouped by issue type, with recommended actions but does not modify anything without explicit confirmation.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3947,3948,3949],{"name":3815,"slug":3816,"type":15},{"name":3850,"slug":3851,"type":15},{"name":9,"slug":8,"type":15},"2026-06-08T08:08:33.693989",{"slug":3952,"name":3952,"fn":3953,"description":3954,"org":3955,"tags":3956,"stars":3938,"repoUrl":3939,"updatedAt":3963},"auditing-warehouse-source-health","audit PostHog data warehouse source health","Audit the health of a PostHog project's data warehouse sources and syncs — find every broken or degraded source connection, sync schema, and webhook channel. Use when the user asks \"why are my imports failing?\", \"what's broken with my sources?\", \"why is my warehouse data stale?\", or wants a one-shot triage of source\u002Fsync health before deciding where to dig in. Produces a prioritized report grouped by severity, with recommended next steps. For materialized-view health use `auditing-warehouse-view-health`; for a single failing sync use `diagnosing-failed-warehouse-syncs`.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3957,3958,3961,3962],{"name":3850,"slug":3851,"type":15},{"name":3959,"slug":3960,"type":15},"Data Warehouse","data-warehouse",{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},"2026-06-18T08:22:57.67984",{"slug":3965,"name":3965,"fn":3966,"description":3967,"org":3968,"tags":3969,"stars":3938,"repoUrl":3939,"updatedAt":3976},"auditing-warehouse-view-health","audit PostHog materialized view health","Audit the health of a PostHog project's materialized views (saved queries) — find every failed materialization and flag unused or stale materialized views that cost storage and compute. Use when the user asks \"which of my views are broken?\", \"why is this materialized view failing?\", \"are any of my views wasting compute?\", or wants a one-shot triage of view health. For source\u002Fsync health use `auditing-warehouse-source-health`.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3970,3971,3972,3975],{"name":3850,"slug":3851,"type":15},{"name":3959,"slug":3960,"type":15},{"name":3973,"slug":3974,"type":15},"Performance","performance",{"name":9,"slug":8,"type":15},"2026-06-18T08:25:10.936787",{"slug":3978,"name":3978,"fn":3979,"description":3980,"org":3981,"tags":3982,"stars":3938,"repoUrl":3939,"updatedAt":3989},"authoring-error-tracking-alerts","author PostHog error tracking alerts","Author error tracking alerts that fire when an issue is created, reopened, or starts spiking. Use when the user asks to set up error notifications, route exceptions to Slack\u002Fwebhook\u002FLinear, or evaluate which error events are worth alerting on. Covers trigger-event selection, integration choice, dedup against existing alerts, and shipping with the canonical message body shape.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3983,3986,3987,3988],{"name":3984,"slug":3985,"type":15},"Alerting","alerting",{"name":27,"slug":28,"type":15},{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},"2026-06-18T08:24:40.318583",{"slug":3991,"name":3991,"fn":3992,"description":3993,"org":3994,"tags":3995,"stars":3938,"repoUrl":3939,"updatedAt":4005},"authoring-log-alerts","author log alerts in PostHog","Author useful, low-noise log alerts on services in a PostHog project. Use when the user asks to set up alerts for their logs, suggest alerts they should add, or evaluate whether a service is worth monitoring. Covers service triage, baseline characterisation, threshold drafting, back-testing via simulate, and shipping with a notification destination.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[3996,3997,4000,4001,4004],{"name":3815,"slug":3816,"type":15},{"name":3998,"slug":3999,"type":15},"Monitoring","monitoring",{"name":17,"slug":18,"type":15},{"name":4002,"slug":4003,"type":15},"Operations","operations",{"name":9,"slug":8,"type":15},"2026-07-18T05:10:54.430898",{"slug":4007,"name":4007,"fn":4008,"description":4009,"org":4010,"tags":4011,"stars":3938,"repoUrl":3939,"updatedAt":4018},"building-workflows","build and edit PostHog workflows","Build, edit, test, enable, and monitor PostHog workflows over MCP. Author the action\u002Fedge graph so it runs and opens cleanly in the visual editor, then change drafts surgically with patch operations. Use when asked to build, set up, automate, change, fix, or debug a workflow, campaign, broadcast, drip sequence, or event-triggered automation in the workflows product.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4012,4013,4014,4015],{"name":3868,"slug":3869,"type":15},{"name":3884,"slug":3885,"type":15},{"name":9,"slug":8,"type":15},{"name":4016,"slug":4017,"type":15},"Workflow Automation","workflow-automation","2026-07-28T05:34:12.167015",{"slug":4020,"name":4020,"fn":4021,"description":4022,"org":4023,"tags":4024,"stars":3938,"repoUrl":3939,"updatedAt":4030},"check-posthog-loading","inspect PostHog SDK loading across URLs","Inspect how the PostHog JavaScript SDK is loaded across a list of URLs. Use to confirm consistent installation across pages, find pages missing the snippet, detect mismatched API keys or hosts between pages, and verify the load method (head snippet vs deferred vs array.js).\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4025,4026,4027,4028,4029],{"name":3815,"slug":3816,"type":15},{"name":27,"slug":28,"type":15},{"name":3833,"slug":3834,"type":15},{"name":17,"slug":18,"type":15},{"name":9,"slug":8,"type":15},"2026-05-07T05:56:19.828048",{"slug":4032,"name":4032,"fn":4033,"description":4034,"org":4035,"tags":4036,"stars":3938,"repoUrl":3939,"updatedAt":4045},"consuming-endpoints-from-client-code","integrate PostHog endpoints into client applications","Wire a PostHog endpoint into a client app or SDK. Covers fetching the OpenAPI spec, generating a typed client with openapi-generator or @hey-api\u002Fopenapi-ts, sending the right auth header, shaping the variables payload (HogQL code_name vs insight breakdown property), handling rate-limit and materialised-endpoint error responses. Use when the user says \"how do I call my endpoint\", \"generate a client for this\", or \"what auth header do I use\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4037,4040,4041,4042],{"name":4038,"slug":4039,"type":15},"API Development","api-development",{"name":3833,"slug":3834,"type":15},{"name":9,"slug":8,"type":15},{"name":4043,"slug":4044,"type":15},"SDK","sdk","2026-06-08T08:08:34.929454",{"slug":4047,"name":4047,"fn":4048,"description":4049,"org":4050,"tags":4051,"stars":3938,"repoUrl":3939,"updatedAt":4055},"copying-endpoints-across-projects","copy PostHog endpoints across projects","Copy a PostHog endpoint (a saved HogQL\u002Finsight query exposed as an API route) to another project in the same organization, or duplicate it under a new name in the same project. Use when the user wants to duplicate an endpoint, promote an endpoint from staging to production, replicate an endpoint's query\u002Fvariables\u002Ffreshness config in another workspace, or clone an endpoint to iterate on it. Unlike feature flags and experiments, endpoints have NO native cross-project copy tool — this skill covers the read-then-recreate flow (endpoint-get then endpoint-create), the active-project switching it requires, name-collision checks, and the safe defaults (land unmaterialised in the target, verify with endpoint-run). Does not cover editing endpoint versions (see managing-endpoint-versions) or authoring a brand-new endpoint from scratch (see creating-an-endpoint).\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4052,4053,4054],{"name":4038,"slug":4039,"type":15},{"name":4002,"slug":4003,"type":15},{"name":9,"slug":8,"type":15},"2026-07-15T05:29:58.442727",{"slug":4057,"name":4057,"fn":4058,"description":4059,"org":4060,"tags":4061,"stars":3938,"repoUrl":3939,"updatedAt":4070},"creating-ai-subscription","schedule recurring AI-generated PostHog reports","Create a recurring AI-generated PostHog report — schedule a free-text prompt to run on a cron, with the LLM-synthesized markdown delivered to email or Slack on each tick. Use when the user wants a recurring AI summary of X on any cadence (daily, weekly, monthly, yearly) rather than a one-off report. (To attach an AI summary to an existing insight\u002Fdashboard subscription instead of a free-text prompt, see `managing-subscriptions` and its `summary_enabled` option.)\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4062,4063,4065,4066,4067],{"name":3868,"slug":3869,"type":15},{"name":4064,"slug":1837,"type":15},"Email",{"name":9,"slug":8,"type":15},{"name":3920,"slug":3921,"type":15},{"name":4068,"slug":4069,"type":15},"Slack","slack","2026-06-09T07:32:27.935712",{"slug":4072,"name":4072,"fn":4073,"description":4074,"org":4075,"tags":4076,"stars":3938,"repoUrl":3939,"updatedAt":4080},"creating-an-endpoint","create PostHog API endpoints","Create a PostHog endpoint with the right shape on the first try — covers query kind choice, name conventions, what to expose as variables (HogQL code_name vs insight breakdown), data_freshness_seconds, and whether to materialise on day one. Use when the user says \"create an endpoint\", \"expose this query as an API\", \"turn this insight into an endpoint\", or asks for help structuring a new endpoint. Steers away from common mistakes: materialising a query with cohort breakdowns or compare mode, inline-only variables on a materialised endpoint, unbounded date ranges, ambiguous names.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[4077,4078,4079],{"name":3815,"slug":3816,"type":15},{"name":4038,"slug":4039,"type":15},{"name":9,"slug":8,"type":15},"2026-06-08T08:08:29.624498",231]