[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-posthog-exploring-mcp-tool-usage":3,"mdc-b14gb1-key":48,"related-repo-posthog-exploring-mcp-tool-usage":896,"related-org-posthog-exploring-mcp-tool-usage":995},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":23,"repoUrl":24,"updatedAt":25,"license":26,"forks":27,"topics":28,"repo":43,"sourceUrl":46,"mdContent":47},"exploring-mcp-tool-usage","explore PostHog MCP tool usage","Starting point for exploring how a PostHog MCP server's tools are used β€” routes a broad question to the typed tool that answers it. Use when the user asks \"how is my MCP doing?\", \"what should I look at?\", \"explore my tool calls\", \"who uses my MCP tools?\", \"what are agents doing with the MCP?\", or pastes an MCP analytics URL without a specific question. Offers a menu of questions, each backed by a query tool, then hands off to the focused skill.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"posthog","PostHog","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fposthog.png",[12,14,17,20],{"name":9,"slug":8,"type":13},"tag",{"name":15,"slug":16,"type":13},"MCP","mcp",{"name":18,"slug":19,"type":13},"Analytics","analytics",{"name":21,"slug":22,"type":13},"Debugging","debugging",35568,"https:\u002F\u002Fgithub.com\u002FPostHog\u002Fposthog","2026-07-18T05:11:01.463824",null,2977,[29,30,19,31,32,33,34,35,36,37,38,39,40,41,42],"ab-testing","ai-analytics","cdp","data-warehouse","experiments","feature-flags","javascript","product-analytics","python","react","session-replay","surveys","typescript","web-analytics",{"repoUrl":24,"stars":23,"forks":27,"topics":44,"description":45},[29,30,19,31,32,33,34,35,36,37,38,39,40,41,42],"πŸ¦” PostHog is an all-in-one developer platform for building successful products. We offer product analytics, web analytics, session replay, error tracking, feature flags, experimentation, surveys, data warehouse, a CDP, and an AI product assistant to help debug your code, ship features faster, and keep all your usage and customer data in one stack.","https:\u002F\u002Fgithub.com\u002FPostHog\u002Fposthog\u002Ftree\u002FHEAD\u002Fproducts\u002Fmcp_analytics\u002Fskills\u002Fexploring-mcp-tool-usage","---\nname: exploring-mcp-tool-usage\ndescription: >\n Starting point for exploring how a PostHog MCP server's tools are used β€”\n routes a broad question to the typed tool that answers it. Use when the user\n asks \"how is my MCP doing?\", \"what should I look at?\", \"explore my tool\n calls\", \"who uses my MCP tools?\", \"what are agents doing with the MCP?\", or\n pastes an MCP analytics URL without a specific question. Offers a menu of\n questions, each backed by a query tool, then hands off to the focused skill.\n---\n\n# Exploring MCP tool usage\n\nAny MCP server instrumented with the `@posthog\u002Fmcp` SDK emits a `$mcp_tool_call`\nevent every time an agent invokes a tool. This skill is the **front door** for a\nuser who knows they want to look at their MCP tool usage but hasn't picked a\nspecific question. Offer the menu below, then route to the tool β€” or the focused\nskill β€” that answers what they choose.\n\nEvery per-tool tool here is gated behind the `mcp-analytics` flag, takes a\n`toolName` (the effective tool name β€” resolved server-side, so pass the name the\nagent actually invokes β€” **except `posthog:query-mcp-tool-failures`**, which\nmatches `$exception` events and so takes the raw registered `$mcp_tool_name`)\nplus a `dateRange`, and runs the same query runner the tool-detail UI uses. So\nresults match the UI, and you never hand-write the HogQL.\n\n## Suggested questions\n\nLead with these when the user is unsure what to ask:\n\n| Ask the user… | Answered by |\n| ------------------------------------------------- | ----------------------------------------------------------------------------------------------- |\n| \"Which tools fail most, or are slowest?\" | `exploring-mcp-tool-quality` (ranks all tools), then `posthog:query-mcp-tool-stats` to drill in |\n| \"How is tool X doing overall?\" | `posthog:query-mcp-tool-stats` β€” calls, errors, p50\u002Fp95, users, sessions, intents |\n| \"How has tool X trended?\" | `posthog:query-mcp-tool-daily-stats` β€” day-by-day series |\n| \"Why is tool X failing?\" | `posthog:query-mcp-tool-failures` β€” top error messages, by harness (raw tool name) |\n| \"Who uses tool X the most?\" | `posthog:query-mcp-tool-top-users` β€” top callers (incl. person email\u002Fname) |\n| \"What gets called right before\u002Fafter tool X?\" | `posthog:query-mcp-tool-neighbors` (`neighborDirection: before`\u002F`after`) |\n| \"What are agents trying to do with tool X?\" | `posthog:query-mcp-tool-sample-intents` β€” recent agent intents |\n| \"What description is tool X registered with?\" | `posthog:query-mcp-tool-descriptions` β€” distinct descriptions seen |\n| \"Which harnesses use my MCP, how reliably?\" | `posthog:query-mcp-harness-breakdown` β€” calls\u002Ferrors\u002Fsessions per client |\n| \"What are agents trying to do, across all tools?\" | `exploring-mcp-intent-clusters` β€” semantic goal clusters |\n| \"Who is connecting, and how active are they?\" | `posthog:mcp-analytics-sessions-list` β€” one row per session, with client and person |\n| \"What did this one session do?\" | `exploring-mcp-sessions` β€” a single agent run's tool sequence |\n\n## Finding the tool name\n\nThe per-tool tools need a `toolName`. If the user named a tool, pass it. If they\nasked a broad \"which tool…\" question, start with `exploring-mcp-tool-quality` to\nrank the tools, pick the one that stands out, then drill in with the per-tool\ntools above. The name to pass is the **effective** tool name (the inner tool for\nsingle-exec wrapper calls) β€” the same string the tool-quality ranking returns.\nThe one exception is `posthog:query-mcp-tool-failures`, which matches `$exception`\nevents by the raw registered `$mcp_tool_name`, not the effective inner tool.\n\n## How to use a per-tool tool\n\nCall it with the tool name and a window, e.g. for the headline numbers of a tool:\n\n```text\nposthog:query-mcp-tool-stats { \"toolName\": \"\u003Ctool>\", \"dateRange\": { \"date_from\": \"-7d\" } }\n```\n\nThen offer a natural follow-up from the menu β€” e.g. after\n`posthog:query-mcp-tool-stats` shows a high error rate, reach for\n`posthog:query-mcp-tool-failures`; after it shows broad reach, reach for\n`posthog:query-mcp-tool-top-users` or `posthog:query-mcp-tool-neighbors`.\n\n## When to drop to SQL\n\n**Covered by a typed tool β€” don't hand-write SQL for these:**\n\n| Question | Tool |\n| ---------------------------------- | ------------------------------------------- |\n| One tool's headline numbers | `posthog:query-mcp-tool-stats` |\n| One tool's day-by-day trend | `posthog:query-mcp-tool-daily-stats` |\n| One tool's top errors | `posthog:query-mcp-tool-failures` |\n| One tool's top callers | `posthog:query-mcp-tool-top-users` |\n| Tools called before\u002Fafter one tool | `posthog:query-mcp-tool-neighbors` |\n| One tool's recent agent intents | `posthog:query-mcp-tool-sample-intents` |\n| One tool's registered descriptions | `posthog:query-mcp-tool-descriptions` |\n| Usage split by client harness | `posthog:query-mcp-harness-breakdown` |\n| List sessions | `posthog:mcp-analytics-sessions-list` |\n| One session's tool calls | `posthog:mcp-analytics-sessions-tool-calls` |\n\n**Not covered β€” use `posthog:execute-sql`:**\n\n- Cross-tool rankings (the tool-quality matrix β€” \"which tool errors most?\")\n- Errored-session filtering (the session list has no error filter or error count)\n- Effective tool names inside a session (`posthog:mcp-analytics-sessions-tool-calls`\n returns the raw `$mcp_tool_name`, not the inner tool of a wrapper call)\n- Any custom breakdown\n\n`posthog:execute-sql` is also the fallback when the `mcp-analytics` flag is off β€”\nevery tool in the table above is gated behind it, `execute-sql` is not. Query\n`$mcp_tool_call` directly; the schema and recipes are in\n[`models-mcp.md`](..\u002F..\u002F..\u002Fposthog_ai\u002Fskills\u002Fquerying-posthog-data\u002Freferences\u002Fmodels-mcp.md).\n\n## Related skills\n\n- [`exploring-mcp-tool-quality`](..\u002Fexploring-mcp-tool-quality\u002FSKILL.md) β€” rank\n tools by error rate \u002F latency \u002F reach, then drill in\n- [`exploring-mcp-sessions`](..\u002Fexploring-mcp-sessions\u002FSKILL.md) β€” a single agent\n run and its tool sequence\n- [`exploring-mcp-intent-clusters`](..\u002Fexploring-mcp-intent-clusters\u002FSKILL.md) β€”\n agent goals grouped by semantic similarity\n",{"data":49,"body":50},{"name":4,"description":6},{"type":51,"children":52},"root",[53,61,92,150,157,162,440,446,493,499,504,516,549,555,563,746,762,801,845,851],{"type":54,"tag":55,"props":56,"children":57},"element","h1",{"id":4},[58],{"type":59,"value":60},"text","Exploring MCP tool usage",{"type":54,"tag":62,"props":63,"children":64},"p",{},[65,67,74,76,82,84,90],{"type":59,"value":66},"Any MCP server instrumented with the ",{"type":54,"tag":68,"props":69,"children":71},"code",{"className":70},[],[72],{"type":59,"value":73},"@posthog\u002Fmcp",{"type":59,"value":75}," SDK emits a ",{"type":54,"tag":68,"props":77,"children":79},{"className":78},[],[80],{"type":59,"value":81},"$mcp_tool_call",{"type":59,"value":83},"\nevent every time an agent invokes a tool. This skill is the ",{"type":54,"tag":85,"props":86,"children":87},"strong",{},[88],{"type":59,"value":89},"front door",{"type":59,"value":91}," for a\nuser who knows they want to look at their MCP tool usage but hasn't picked a\nspecific question. Offer the menu below, then route to the tool β€” or the focused\nskill β€” that answers what they choose.",{"type":54,"tag":62,"props":93,"children":94},{},[95,97,103,105,111,113,124,126,132,134,140,142,148],{"type":59,"value":96},"Every per-tool tool here is gated behind the ",{"type":54,"tag":68,"props":98,"children":100},{"className":99},[],[101],{"type":59,"value":102},"mcp-analytics",{"type":59,"value":104}," flag, takes a\n",{"type":54,"tag":68,"props":106,"children":108},{"className":107},[],[109],{"type":59,"value":110},"toolName",{"type":59,"value":112}," (the effective tool name β€” resolved server-side, so pass the name the\nagent actually invokes β€” ",{"type":54,"tag":85,"props":114,"children":115},{},[116,118],{"type":59,"value":117},"except ",{"type":54,"tag":68,"props":119,"children":121},{"className":120},[],[122],{"type":59,"value":123},"posthog:query-mcp-tool-failures",{"type":59,"value":125},", which\nmatches ",{"type":54,"tag":68,"props":127,"children":129},{"className":128},[],[130],{"type":59,"value":131},"$exception",{"type":59,"value":133}," events and so takes the raw registered ",{"type":54,"tag":68,"props":135,"children":137},{"className":136},[],[138],{"type":59,"value":139},"$mcp_tool_name",{"type":59,"value":141},")\nplus a ",{"type":54,"tag":68,"props":143,"children":145},{"className":144},[],[146],{"type":59,"value":147},"dateRange",{"type":59,"value":149},", and runs the same query runner the tool-detail UI uses. So\nresults match the UI, and you never hand-write the HogQL.",{"type":54,"tag":151,"props":152,"children":154},"h2",{"id":153},"suggested-questions",[155],{"type":59,"value":156},"Suggested questions",{"type":54,"tag":62,"props":158,"children":159},{},[160],{"type":59,"value":161},"Lead with these when the user is unsure what to ask:",{"type":54,"tag":163,"props":164,"children":165},"table",{},[166,185],{"type":54,"tag":167,"props":168,"children":169},"thead",{},[170],{"type":54,"tag":171,"props":172,"children":173},"tr",{},[174,180],{"type":54,"tag":175,"props":176,"children":177},"th",{},[178],{"type":59,"value":179},"Ask the user…",{"type":54,"tag":175,"props":181,"children":182},{},[183],{"type":59,"value":184},"Answered by",{"type":54,"tag":186,"props":187,"children":188},"tbody",{},[189,217,235,254,272,291,326,345,364,383,402,421],{"type":54,"tag":171,"props":190,"children":191},{},[192,198],{"type":54,"tag":193,"props":194,"children":195},"td",{},[196],{"type":59,"value":197},"\"Which tools fail most, or are slowest?\"",{"type":54,"tag":193,"props":199,"children":200},{},[201,207,209,215],{"type":54,"tag":68,"props":202,"children":204},{"className":203},[],[205],{"type":59,"value":206},"exploring-mcp-tool-quality",{"type":59,"value":208}," (ranks all tools), then ",{"type":54,"tag":68,"props":210,"children":212},{"className":211},[],[213],{"type":59,"value":214},"posthog:query-mcp-tool-stats",{"type":59,"value":216}," to drill in",{"type":54,"tag":171,"props":218,"children":219},{},[220,225],{"type":54,"tag":193,"props":221,"children":222},{},[223],{"type":59,"value":224},"\"How is tool X doing overall?\"",{"type":54,"tag":193,"props":226,"children":227},{},[228,233],{"type":54,"tag":68,"props":229,"children":231},{"className":230},[],[232],{"type":59,"value":214},{"type":59,"value":234}," β€” calls, errors, p50\u002Fp95, users, sessions, intents",{"type":54,"tag":171,"props":236,"children":237},{},[238,243],{"type":54,"tag":193,"props":239,"children":240},{},[241],{"type":59,"value":242},"\"How has tool X trended?\"",{"type":54,"tag":193,"props":244,"children":245},{},[246,252],{"type":54,"tag":68,"props":247,"children":249},{"className":248},[],[250],{"type":59,"value":251},"posthog:query-mcp-tool-daily-stats",{"type":59,"value":253}," β€” day-by-day series",{"type":54,"tag":171,"props":255,"children":256},{},[257,262],{"type":54,"tag":193,"props":258,"children":259},{},[260],{"type":59,"value":261},"\"Why is tool X failing?\"",{"type":54,"tag":193,"props":263,"children":264},{},[265,270],{"type":54,"tag":68,"props":266,"children":268},{"className":267},[],[269],{"type":59,"value":123},{"type":59,"value":271}," β€” top error messages, by harness (raw tool name)",{"type":54,"tag":171,"props":273,"children":274},{},[275,280],{"type":54,"tag":193,"props":276,"children":277},{},[278],{"type":59,"value":279},"\"Who uses tool X the most?\"",{"type":54,"tag":193,"props":281,"children":282},{},[283,289],{"type":54,"tag":68,"props":284,"children":286},{"className":285},[],[287],{"type":59,"value":288},"posthog:query-mcp-tool-top-users",{"type":59,"value":290}," β€” top callers (incl. person email\u002Fname)",{"type":54,"tag":171,"props":292,"children":293},{},[294,299],{"type":54,"tag":193,"props":295,"children":296},{},[297],{"type":59,"value":298},"\"What gets called right before\u002Fafter tool X?\"",{"type":54,"tag":193,"props":300,"children":301},{},[302,308,310,316,318,324],{"type":54,"tag":68,"props":303,"children":305},{"className":304},[],[306],{"type":59,"value":307},"posthog:query-mcp-tool-neighbors",{"type":59,"value":309}," (",{"type":54,"tag":68,"props":311,"children":313},{"className":312},[],[314],{"type":59,"value":315},"neighborDirection: before",{"type":59,"value":317},"\u002F",{"type":54,"tag":68,"props":319,"children":321},{"className":320},[],[322],{"type":59,"value":323},"after",{"type":59,"value":325},")",{"type":54,"tag":171,"props":327,"children":328},{},[329,334],{"type":54,"tag":193,"props":330,"children":331},{},[332],{"type":59,"value":333},"\"What are agents trying to do with tool X?\"",{"type":54,"tag":193,"props":335,"children":336},{},[337,343],{"type":54,"tag":68,"props":338,"children":340},{"className":339},[],[341],{"type":59,"value":342},"posthog:query-mcp-tool-sample-intents",{"type":59,"value":344}," β€” recent agent intents",{"type":54,"tag":171,"props":346,"children":347},{},[348,353],{"type":54,"tag":193,"props":349,"children":350},{},[351],{"type":59,"value":352},"\"What description is tool X registered with?\"",{"type":54,"tag":193,"props":354,"children":355},{},[356,362],{"type":54,"tag":68,"props":357,"children":359},{"className":358},[],[360],{"type":59,"value":361},"posthog:query-mcp-tool-descriptions",{"type":59,"value":363}," β€” distinct descriptions seen",{"type":54,"tag":171,"props":365,"children":366},{},[367,372],{"type":54,"tag":193,"props":368,"children":369},{},[370],{"type":59,"value":371},"\"Which harnesses use my MCP, how reliably?\"",{"type":54,"tag":193,"props":373,"children":374},{},[375,381],{"type":54,"tag":68,"props":376,"children":378},{"className":377},[],[379],{"type":59,"value":380},"posthog:query-mcp-harness-breakdown",{"type":59,"value":382}," β€” calls\u002Ferrors\u002Fsessions per client",{"type":54,"tag":171,"props":384,"children":385},{},[386,391],{"type":54,"tag":193,"props":387,"children":388},{},[389],{"type":59,"value":390},"\"What are agents trying to do, across all tools?\"",{"type":54,"tag":193,"props":392,"children":393},{},[394,400],{"type":54,"tag":68,"props":395,"children":397},{"className":396},[],[398],{"type":59,"value":399},"exploring-mcp-intent-clusters",{"type":59,"value":401}," β€” semantic goal clusters",{"type":54,"tag":171,"props":403,"children":404},{},[405,410],{"type":54,"tag":193,"props":406,"children":407},{},[408],{"type":59,"value":409},"\"Who is connecting, and how active are they?\"",{"type":54,"tag":193,"props":411,"children":412},{},[413,419],{"type":54,"tag":68,"props":414,"children":416},{"className":415},[],[417],{"type":59,"value":418},"posthog:mcp-analytics-sessions-list",{"type":59,"value":420}," β€” one row per session, with client and person",{"type":54,"tag":171,"props":422,"children":423},{},[424,429],{"type":54,"tag":193,"props":425,"children":426},{},[427],{"type":59,"value":428},"\"What did this one session do?\"",{"type":54,"tag":193,"props":430,"children":431},{},[432,438],{"type":54,"tag":68,"props":433,"children":435},{"className":434},[],[436],{"type":59,"value":437},"exploring-mcp-sessions",{"type":59,"value":439}," β€” a single agent run's tool sequence",{"type":54,"tag":151,"props":441,"children":443},{"id":442},"finding-the-tool-name",[444],{"type":59,"value":445},"Finding the tool name",{"type":54,"tag":62,"props":447,"children":448},{},[449,451,456,458,463,465,470,472,477,479,484,486,491],{"type":59,"value":450},"The per-tool tools need a ",{"type":54,"tag":68,"props":452,"children":454},{"className":453},[],[455],{"type":59,"value":110},{"type":59,"value":457},". If the user named a tool, pass it. If they\nasked a broad \"which tool…\" question, start with ",{"type":54,"tag":68,"props":459,"children":461},{"className":460},[],[462],{"type":59,"value":206},{"type":59,"value":464}," to\nrank the tools, pick the one that stands out, then drill in with the per-tool\ntools above. The name to pass is the ",{"type":54,"tag":85,"props":466,"children":467},{},[468],{"type":59,"value":469},"effective",{"type":59,"value":471}," tool name (the inner tool for\nsingle-exec wrapper calls) β€” the same string the tool-quality ranking returns.\nThe one exception is ",{"type":54,"tag":68,"props":473,"children":475},{"className":474},[],[476],{"type":59,"value":123},{"type":59,"value":478},", which matches ",{"type":54,"tag":68,"props":480,"children":482},{"className":481},[],[483],{"type":59,"value":131},{"type":59,"value":485},"\nevents by the raw registered ",{"type":54,"tag":68,"props":487,"children":489},{"className":488},[],[490],{"type":59,"value":139},{"type":59,"value":492},", not the effective inner tool.",{"type":54,"tag":151,"props":494,"children":496},{"id":495},"how-to-use-a-per-tool-tool",[497],{"type":59,"value":498},"How to use a per-tool tool",{"type":54,"tag":62,"props":500,"children":501},{},[502],{"type":59,"value":503},"Call it with the tool name and a window, e.g. for the headline numbers of a tool:",{"type":54,"tag":505,"props":506,"children":511},"pre",{"className":507,"code":509,"language":59,"meta":510},[508],"language-text","posthog:query-mcp-tool-stats { \"toolName\": \"\u003Ctool>\", \"dateRange\": { \"date_from\": \"-7d\" } }\n","",[512],{"type":54,"tag":68,"props":513,"children":514},{"__ignoreMap":510},[515],{"type":59,"value":509},{"type":54,"tag":62,"props":517,"children":518},{},[519,521,526,528,533,535,540,542,547],{"type":59,"value":520},"Then offer a natural follow-up from the menu β€” e.g. after\n",{"type":54,"tag":68,"props":522,"children":524},{"className":523},[],[525],{"type":59,"value":214},{"type":59,"value":527}," shows a high error rate, reach for\n",{"type":54,"tag":68,"props":529,"children":531},{"className":530},[],[532],{"type":59,"value":123},{"type":59,"value":534},"; after it shows broad reach, reach for\n",{"type":54,"tag":68,"props":536,"children":538},{"className":537},[],[539],{"type":59,"value":288},{"type":59,"value":541}," or ",{"type":54,"tag":68,"props":543,"children":545},{"className":544},[],[546],{"type":59,"value":307},{"type":59,"value":548},".",{"type":54,"tag":151,"props":550,"children":552},{"id":551},"when-to-drop-to-sql",[553],{"type":59,"value":554},"When to drop to SQL",{"type":54,"tag":62,"props":556,"children":557},{},[558],{"type":54,"tag":85,"props":559,"children":560},{},[561],{"type":59,"value":562},"Covered by a typed tool β€” don't hand-write SQL for these:",{"type":54,"tag":163,"props":564,"children":565},{},[566,582],{"type":54,"tag":167,"props":567,"children":568},{},[569],{"type":54,"tag":171,"props":570,"children":571},{},[572,577],{"type":54,"tag":175,"props":573,"children":574},{},[575],{"type":59,"value":576},"Question",{"type":54,"tag":175,"props":578,"children":579},{},[580],{"type":59,"value":581},"Tool",{"type":54,"tag":186,"props":583,"children":584},{},[585,601,617,633,649,665,681,697,713,729],{"type":54,"tag":171,"props":586,"children":587},{},[588,593],{"type":54,"tag":193,"props":589,"children":590},{},[591],{"type":59,"value":592},"One tool's headline numbers",{"type":54,"tag":193,"props":594,"children":595},{},[596],{"type":54,"tag":68,"props":597,"children":599},{"className":598},[],[600],{"type":59,"value":214},{"type":54,"tag":171,"props":602,"children":603},{},[604,609],{"type":54,"tag":193,"props":605,"children":606},{},[607],{"type":59,"value":608},"One tool's day-by-day trend",{"type":54,"tag":193,"props":610,"children":611},{},[612],{"type":54,"tag":68,"props":613,"children":615},{"className":614},[],[616],{"type":59,"value":251},{"type":54,"tag":171,"props":618,"children":619},{},[620,625],{"type":54,"tag":193,"props":621,"children":622},{},[623],{"type":59,"value":624},"One tool's top errors",{"type":54,"tag":193,"props":626,"children":627},{},[628],{"type":54,"tag":68,"props":629,"children":631},{"className":630},[],[632],{"type":59,"value":123},{"type":54,"tag":171,"props":634,"children":635},{},[636,641],{"type":54,"tag":193,"props":637,"children":638},{},[639],{"type":59,"value":640},"One tool's top callers",{"type":54,"tag":193,"props":642,"children":643},{},[644],{"type":54,"tag":68,"props":645,"children":647},{"className":646},[],[648],{"type":59,"value":288},{"type":54,"tag":171,"props":650,"children":651},{},[652,657],{"type":54,"tag":193,"props":653,"children":654},{},[655],{"type":59,"value":656},"Tools called before\u002Fafter one tool",{"type":54,"tag":193,"props":658,"children":659},{},[660],{"type":54,"tag":68,"props":661,"children":663},{"className":662},[],[664],{"type":59,"value":307},{"type":54,"tag":171,"props":666,"children":667},{},[668,673],{"type":54,"tag":193,"props":669,"children":670},{},[671],{"type":59,"value":672},"One tool's recent agent intents",{"type":54,"tag":193,"props":674,"children":675},{},[676],{"type":54,"tag":68,"props":677,"children":679},{"className":678},[],[680],{"type":59,"value":342},{"type":54,"tag":171,"props":682,"children":683},{},[684,689],{"type":54,"tag":193,"props":685,"children":686},{},[687],{"type":59,"value":688},"One tool's registered descriptions",{"type":54,"tag":193,"props":690,"children":691},{},[692],{"type":54,"tag":68,"props":693,"children":695},{"className":694},[],[696],{"type":59,"value":361},{"type":54,"tag":171,"props":698,"children":699},{},[700,705],{"type":54,"tag":193,"props":701,"children":702},{},[703],{"type":59,"value":704},"Usage split by client harness",{"type":54,"tag":193,"props":706,"children":707},{},[708],{"type":54,"tag":68,"props":709,"children":711},{"className":710},[],[712],{"type":59,"value":380},{"type":54,"tag":171,"props":714,"children":715},{},[716,721],{"type":54,"tag":193,"props":717,"children":718},{},[719],{"type":59,"value":720},"List sessions",{"type":54,"tag":193,"props":722,"children":723},{},[724],{"type":54,"tag":68,"props":725,"children":727},{"className":726},[],[728],{"type":59,"value":418},{"type":54,"tag":171,"props":730,"children":731},{},[732,737],{"type":54,"tag":193,"props":733,"children":734},{},[735],{"type":59,"value":736},"One session's tool calls",{"type":54,"tag":193,"props":738,"children":739},{},[740],{"type":54,"tag":68,"props":741,"children":743},{"className":742},[],[744],{"type":59,"value":745},"posthog:mcp-analytics-sessions-tool-calls",{"type":54,"tag":62,"props":747,"children":748},{},[749],{"type":54,"tag":85,"props":750,"children":751},{},[752,754,760],{"type":59,"value":753},"Not covered β€” use ",{"type":54,"tag":68,"props":755,"children":757},{"className":756},[],[758],{"type":59,"value":759},"posthog:execute-sql",{"type":59,"value":761},":",{"type":54,"tag":763,"props":764,"children":765},"ul",{},[766,772,777,796],{"type":54,"tag":767,"props":768,"children":769},"li",{},[770],{"type":59,"value":771},"Cross-tool rankings (the tool-quality matrix β€” \"which tool errors most?\")",{"type":54,"tag":767,"props":773,"children":774},{},[775],{"type":59,"value":776},"Errored-session filtering (the session list has no error filter or error count)",{"type":54,"tag":767,"props":778,"children":779},{},[780,782,787,789,794],{"type":59,"value":781},"Effective tool names inside a session (",{"type":54,"tag":68,"props":783,"children":785},{"className":784},[],[786],{"type":59,"value":745},{"type":59,"value":788},"\nreturns the raw ",{"type":54,"tag":68,"props":790,"children":792},{"className":791},[],[793],{"type":59,"value":139},{"type":59,"value":795},", not the inner tool of a wrapper call)",{"type":54,"tag":767,"props":797,"children":798},{},[799],{"type":59,"value":800},"Any custom breakdown",{"type":54,"tag":62,"props":802,"children":803},{},[804,809,811,816,818,824,826,831,833,844],{"type":54,"tag":68,"props":805,"children":807},{"className":806},[],[808],{"type":59,"value":759},{"type":59,"value":810}," is also the fallback when the ",{"type":54,"tag":68,"props":812,"children":814},{"className":813},[],[815],{"type":59,"value":102},{"type":59,"value":817}," flag is off β€”\nevery tool in the table above is gated behind it, ",{"type":54,"tag":68,"props":819,"children":821},{"className":820},[],[822],{"type":59,"value":823},"execute-sql",{"type":59,"value":825}," is not. Query\n",{"type":54,"tag":68,"props":827,"children":829},{"className":828},[],[830],{"type":59,"value":81},{"type":59,"value":832}," directly; the schema and recipes are in\n",{"type":54,"tag":834,"props":835,"children":837},"a",{"href":836},"..\u002F..\u002F..\u002Fposthog_ai\u002Fskills\u002Fquerying-posthog-data\u002Freferences\u002Fmodels-mcp.md",[838],{"type":54,"tag":68,"props":839,"children":841},{"className":840},[],[842],{"type":59,"value":843},"models-mcp.md",{"type":59,"value":548},{"type":54,"tag":151,"props":846,"children":848},{"id":847},"related-skills",[849],{"type":59,"value":850},"Related skills",{"type":54,"tag":763,"props":852,"children":853},{},[854,868,882],{"type":54,"tag":767,"props":855,"children":856},{},[857,866],{"type":54,"tag":834,"props":858,"children":860},{"href":859},"..\u002Fexploring-mcp-tool-quality\u002FSKILL.md",[861],{"type":54,"tag":68,"props":862,"children":864},{"className":863},[],[865],{"type":59,"value":206},{"type":59,"value":867}," β€” rank\ntools by error rate \u002F latency \u002F reach, then drill in",{"type":54,"tag":767,"props":869,"children":870},{},[871,880],{"type":54,"tag":834,"props":872,"children":874},{"href":873},"..\u002Fexploring-mcp-sessions\u002FSKILL.md",[875],{"type":54,"tag":68,"props":876,"children":878},{"className":877},[],[879],{"type":59,"value":437},{"type":59,"value":881}," β€” a single agent\nrun and its tool sequence",{"type":54,"tag":767,"props":883,"children":884},{},[885,894],{"type":54,"tag":834,"props":886,"children":888},{"href":887},"..\u002Fexploring-mcp-intent-clusters\u002FSKILL.md",[889],{"type":54,"tag":68,"props":890,"children":892},{"className":891},[],[893],{"type":59,"value":399},{"type":59,"value":895}," β€”\nagent goals grouped by semantic similarity",{"items":897,"total":994},[898,913,925,937,950,963,979],{"slug":899,"name":899,"fn":900,"description":901,"org":902,"tags":903,"stars":23,"repoUrl":24,"updatedAt":912},"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},[904,905,908,911],{"name":18,"slug":19,"type":13},{"name":906,"slug":907,"type":13},"Cost Optimization","cost-optimization",{"name":909,"slug":910,"type":13},"Observability","observability",{"name":9,"slug":8,"type":13},"2026-07-28T05:34:11.117757",{"slug":914,"name":914,"fn":915,"description":916,"org":917,"tags":918,"stars":23,"repoUrl":24,"updatedAt":924},"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},[919,920,923],{"name":18,"slug":19,"type":13},{"name":921,"slug":922,"type":13},"Audit","audit",{"name":9,"slug":8,"type":13},"2026-06-08T08:08:33.693989",{"slug":926,"name":926,"fn":927,"description":928,"org":929,"tags":930,"stars":23,"repoUrl":24,"updatedAt":936},"auditing-warehouse-source-health","audit PostHog data warehouse source health","Audit the health of a PostHog project's data warehouse sources and syncs β€” 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},[931,932,934,935],{"name":921,"slug":922,"type":13},{"name":933,"slug":32,"type":13},"Data Warehouse",{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},"2026-06-18T08:22:57.67984",{"slug":938,"name":938,"fn":939,"description":940,"org":941,"tags":942,"stars":23,"repoUrl":24,"updatedAt":949},"auditing-warehouse-view-health","audit PostHog materialized view health","Audit the health of a PostHog project's materialized views (saved queries) β€” 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},[943,944,945,948],{"name":921,"slug":922,"type":13},{"name":933,"slug":32,"type":13},{"name":946,"slug":947,"type":13},"Performance","performance",{"name":9,"slug":8,"type":13},"2026-06-18T08:25:10.936787",{"slug":951,"name":951,"fn":952,"description":953,"org":954,"tags":955,"stars":23,"repoUrl":24,"updatedAt":962},"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},[956,959,960,961],{"name":957,"slug":958,"type":13},"Alerting","alerting",{"name":21,"slug":22,"type":13},{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},"2026-06-18T08:24:40.318583",{"slug":964,"name":964,"fn":965,"description":966,"org":967,"tags":968,"stars":23,"repoUrl":24,"updatedAt":978},"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},[969,970,973,974,977],{"name":18,"slug":19,"type":13},{"name":971,"slug":972,"type":13},"Monitoring","monitoring",{"name":909,"slug":910,"type":13},{"name":975,"slug":976,"type":13},"Operations","operations",{"name":9,"slug":8,"type":13},"2026-07-18T05:10:54.430898",{"slug":980,"name":980,"fn":981,"description":982,"org":983,"tags":984,"stars":23,"repoUrl":24,"updatedAt":993},"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},[985,988,989,990],{"name":986,"slug":987,"type":13},"Automation","automation",{"name":15,"slug":16,"type":13},{"name":9,"slug":8,"type":13},{"name":991,"slug":992,"type":13},"Workflow Automation","workflow-automation","2026-07-28T05:34:12.167015",61,{"items":996,"total":1113},[997,1004,1010,1017,1024,1031,1039,1046,1060,1075,1085,1103],{"slug":899,"name":899,"fn":900,"description":901,"org":998,"tags":999,"stars":23,"repoUrl":24,"updatedAt":912},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1000,1001,1002,1003],{"name":18,"slug":19,"type":13},{"name":906,"slug":907,"type":13},{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},{"slug":914,"name":914,"fn":915,"description":916,"org":1005,"tags":1006,"stars":23,"repoUrl":24,"updatedAt":924},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1007,1008,1009],{"name":18,"slug":19,"type":13},{"name":921,"slug":922,"type":13},{"name":9,"slug":8,"type":13},{"slug":926,"name":926,"fn":927,"description":928,"org":1011,"tags":1012,"stars":23,"repoUrl":24,"updatedAt":936},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1013,1014,1015,1016],{"name":921,"slug":922,"type":13},{"name":933,"slug":32,"type":13},{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},{"slug":938,"name":938,"fn":939,"description":940,"org":1018,"tags":1019,"stars":23,"repoUrl":24,"updatedAt":949},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1020,1021,1022,1023],{"name":921,"slug":922,"type":13},{"name":933,"slug":32,"type":13},{"name":946,"slug":947,"type":13},{"name":9,"slug":8,"type":13},{"slug":951,"name":951,"fn":952,"description":953,"org":1025,"tags":1026,"stars":23,"repoUrl":24,"updatedAt":962},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1027,1028,1029,1030],{"name":957,"slug":958,"type":13},{"name":21,"slug":22,"type":13},{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},{"slug":964,"name":964,"fn":965,"description":966,"org":1032,"tags":1033,"stars":23,"repoUrl":24,"updatedAt":978},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1034,1035,1036,1037,1038],{"name":18,"slug":19,"type":13},{"name":971,"slug":972,"type":13},{"name":909,"slug":910,"type":13},{"name":975,"slug":976,"type":13},{"name":9,"slug":8,"type":13},{"slug":980,"name":980,"fn":981,"description":982,"org":1040,"tags":1041,"stars":23,"repoUrl":24,"updatedAt":993},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1042,1043,1044,1045],{"name":986,"slug":987,"type":13},{"name":15,"slug":16,"type":13},{"name":9,"slug":8,"type":13},{"name":991,"slug":992,"type":13},{"slug":1047,"name":1047,"fn":1048,"description":1049,"org":1050,"tags":1051,"stars":23,"repoUrl":24,"updatedAt":1059},"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},[1052,1053,1054,1057,1058],{"name":18,"slug":19,"type":13},{"name":21,"slug":22,"type":13},{"name":1055,"slug":1056,"type":13},"Frontend","frontend",{"name":909,"slug":910,"type":13},{"name":9,"slug":8,"type":13},"2026-05-07T05:56:19.828048",{"slug":1061,"name":1061,"fn":1062,"description":1063,"org":1064,"tags":1065,"stars":23,"repoUrl":24,"updatedAt":1074},"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},[1066,1069,1070,1071],{"name":1067,"slug":1068,"type":13},"API Development","api-development",{"name":1055,"slug":1056,"type":13},{"name":9,"slug":8,"type":13},{"name":1072,"slug":1073,"type":13},"SDK","sdk","2026-06-08T08:08:34.929454",{"slug":1076,"name":1076,"fn":1077,"description":1078,"org":1079,"tags":1080,"stars":23,"repoUrl":24,"updatedAt":1084},"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 β€” 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},[1081,1082,1083],{"name":1067,"slug":1068,"type":13},{"name":975,"slug":976,"type":13},{"name":9,"slug":8,"type":13},"2026-07-15T05:29:58.442727",{"slug":1086,"name":1086,"fn":1087,"description":1088,"org":1089,"tags":1090,"stars":23,"repoUrl":24,"updatedAt":1102},"creating-ai-subscription","schedule recurring AI-generated PostHog reports","Create a recurring AI-generated PostHog report β€” 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},[1091,1092,1095,1096,1099],{"name":986,"slug":987,"type":13},{"name":1093,"slug":1094,"type":13},"Email","email",{"name":9,"slug":8,"type":13},{"name":1097,"slug":1098,"type":13},"Reporting","reporting",{"name":1100,"slug":1101,"type":13},"Slack","slack","2026-06-09T07:32:27.935712",{"slug":1104,"name":1104,"fn":1105,"description":1106,"org":1107,"tags":1108,"stars":23,"repoUrl":24,"updatedAt":1112},"creating-an-endpoint","create PostHog API endpoints","Create a PostHog endpoint with the right shape on the first try β€” 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},[1109,1110,1111],{"name":18,"slug":19,"type":13},{"name":1067,"slug":1068,"type":13},{"name":9,"slug":8,"type":13},"2026-06-08T08:08:29.624498",231]