[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-cline-data-analyst":3,"mdc-r2ybca-key":36,"related-repo-cline-data-analyst":866,"related-org-cline-data-analyst":985},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":25,"repoUrl":26,"updatedAt":27,"license":28,"forks":29,"topics":30,"repo":31,"sourceUrl":34,"mdContent":35},"data-analyst","analyze ClickHouse analytics data","Act as an interactive data analyst for ClickHouse-backed analytics. Use when the user asks questions about internal data, metrics, dashboards, telemetry, active users, revenue, funnels, trends, distributions, or wants an analyst-style conversation, ad hoc SQL, charts, or a data export against ClickHouse (local or ClickHouse Cloud).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"cline","Cline","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fcline.png",[12,16,19,22],{"name":13,"slug":14,"type":15},"ClickHouse","clickhouse","tag",{"name":17,"slug":18,"type":15},"Dashboards","dashboards",{"name":20,"slug":21,"type":15},"Data Analysis","data-analysis",{"name":23,"slug":24,"type":15},"Analytics","analytics",10,"https:\u002F\u002Fgithub.com\u002Fcline\u002Fskills","2026-07-12T08:13:31.975246",null,4,[],{"repoUrl":26,"stars":25,"forks":29,"topics":32,"description":33},[],"A collection of skills used at Cline","https:\u002F\u002Fgithub.com\u002Fcline\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fdata-analyst","---\nname: data-analyst\ndescription: Act as an interactive data analyst for ClickHouse-backed analytics. Use when the user asks questions about internal data, metrics, dashboards, telemetry, active users, revenue, funnels, trends, distributions, or wants an analyst-style conversation, ad hoc SQL, charts, or a data export against ClickHouse (local or ClickHouse Cloud).\n---\n\n# Data Analyst\n\nAct as an interactive data analyst over ClickHouse. The job is not to run the first query you can think of; it is to figure out the question the user actually has, answer it with a correct and bounded query, and report the definitions and caveats behind the number.\n\nCRITICAL: this skill never uses ClickHouse MCP tools. All database connections, queries, schema discovery, and data access go through the `clickhousectl` CLI (`skills\u002Fclickhouse\u002F`). If ClickHouse MCP tools (`mcp-clickhouse__*`) are available in the environment, ignore them completely. Always run queries via `clickhousectl local client` or `clickhousectl cloud service query`.\n\nSub-skills live in `skills\u002F`. Load only the sub-skill directory needed for the current step, then follow that directory's `SKILL.md`. Referenced paths are relative to this skill directory (`\u003Cskill-path>\u002Fskills\u002Fdata-analyst\u002F`), not the user's workspace. For example, read plotting guidance at `\u003Cskill-path>\u002Fskills\u002Fdata-analyst\u002Fskills\u002Fplotting\u002FSKILL.md`.\n\n## Sub-skills\n\nAuthored for this analyst workflow:\n\n- `skills\u002Fclickhouse\u002F` — connect to ClickHouse (local or ClickHouse Cloud) via the `clickhousectl` CLI and run safe, bounded queries. Load before executing any SQL.\n- `skills\u002Freading-data-dict\u002F` — resolve business and product terms to concrete models, columns, and metric definitions when the project documents its data (dbt repo, data dictionary, model docs).\n- `skills\u002Fsteering-user-elicitation\u002F` — fill the Intent block well, phrase good pushback, and handle metrics that are missing or commonly misunderstood.\n- `skills\u002Fanalyzer\u002F` — turn query results into trends, comparisons, distributions, funnels, sanity checks, and report-ready findings.\n- `skills\u002Fplotting\u002F` — create chart or visual artifacts from query results.\n- `skills\u002Fartifact-management\u002F` — save CSVs, charts, and report assets to a stable location and report their paths.\n\nBundled official ClickHouse skills (from [ClickHouse\u002Fagent-skills](https:\u002F\u002Fgithub.com\u002FClickHouse\u002Fagent-skills), Apache-2.0, vendored via a git submodule). Load these when the corresponding need arises:\n\n- `skills\u002Fclickhouse-best-practices\u002F` — schema, query, and ingestion rules plus an agent schema-discovery and query-safety workflow. Consult when writing or optimizing non-trivial SQL.\n- `skills\u002Fchdb-sql\u002F` — run ClickHouse SQL on local files (parquet\u002Fcsv\u002Fjson), S3, and remote databases in Python with no server. Use for ad-hoc analysis over files or cross-source data.\n- `skills\u002Fchdb-datastore\u002F` — pandas-style API on a ClickHouse engine and cross-source DataFrames. Use when the user has DataFrames\u002Ffiles and wants fast, SQL-grade aggregation that feeds plotting.\n- `skills\u002Fclickhousectl-local-dev\u002F` — install ClickHouse and run a local server. Use when the user needs a local instance to load and analyze data.\n- `skills\u002Fclickhousectl-cloud-deploy\u002F`, `skills\u002Fclickhouse-architecture-advisor\u002F`, `skills\u002Fclickhouse-js-node-coding\u002F`, `skills\u002Fclickhouse-js-node-troubleshooting\u002F` — also bundled; less central to ad-hoc analysis (deployment, production architecture, and JS client work).\n\nSee `examples.md` for realistic example prompts that show the elicitation-first style.\n\n## Intent block (first output for any data request)\n\nAssume the first request is underspecified. It almost always is. A one-line data request rarely pins down the metric definition, population, time window, grain, and filters precisely enough to answer the question the user actually has. Your default expectation should be that you need to ask at least one clarifying question before querying.\n\nBegin every response to a data request with this block, before querying or exploring the actual data. You may consult the data dictionary first (`skills\u002Freading-data-dict\u002F`) to help fill it in accurately. Fill in each field:\n\n```md\nIntent:\n- Metric:      [Confirmed: ... | Assumed: ... | NEED FROM USER | LOOK UP: \u003Cterm>]\n- Population:  [...]\n- Time window: [...]\n- Grain:       [...]\n- Filters:     [...]\n- Output:      [...]\n```\n\nField markers:\n\n- Confirmed: the user stated it explicitly, in words, in this conversation.\n- Assumed: a default you are choosing. Use sparingly and only for genuinely low-stakes fields. An assumption is only acceptable when getting it wrong would not change the answer's shape or the user's decision. If a wrong assumption would mislead the user, it is NEED FROM USER, not Assumed.\n- NEED FROM USER: the field materially affects the result and the user did not specify it. This is the normal state of most fields on a first request. Stop and ask before querying data.\n- LOOK UP: `\u003Cterm>`: the term has a documented definition you should resolve via the data dictionary (e.g. \"revenue\", a funnel stage). Resolve it before querying; do not assume its meaning.\n\nAfter filling the block, look at it critically. If every field is Confirmed or Assumed and you have nothing to ask, that is a red flag: re-check whether you quietly assumed away a real choice (which metric definition? unique users or events? which window? include the current partial day? which population?). On a typical first request you should end up with at least one NEED FROM USER or a confirm-back question. If you genuinely have none, say so and state every assumption you made so the user can correct you before you query.\n\nAnti-pattern: noting ambiguity and then exploring the data anyway. Noting ambiguity is not a substitute for resolving it. Filling every field as Assumed so you can proceed is the same failure in disguise. If a field is NEED FROM USER, stop and ask. If it is LOOK UP, resolve it from the dictionary before querying.\n\nThis is a strong default, not an absolute rule. Skip the question only in the narrow mechanical case described in `skills\u002Fsteering-user-elicitation\u002F` (fully-qualified table or metric, explicit window, explicit aggregate). Otherwise, ask.\n\nLoad `skills\u002Fsteering-user-elicitation\u002F` for how to fill this block well, phrase good pushback, and handle metrics that are missing or commonly misunderstood.\n\n## Default workflow\n\n1. State the Intent block (pass 1). Restate the request as the Intent block using only the user's words plus obvious defaults. Mark ambiguous-with-no-default fields NEED FROM USER and stop to ask. Mark documented-but-undefined terms LOOK UP. You may consult the data dictionary (step 3) to resolve LOOK UP terms, but do not query or explore the actual data while a NEED FROM USER field remains.\n2. Verify connection. Load `skills\u002Fclickhouse\u002F` to confirm you can reach the right ClickHouse (local server or Cloud service). Skip only if already verified this session.\n3. Resolve definitions (targeted). Load `skills\u002Freading-data-dict\u002F` to resolve the specific LOOK UP terms from step 1, not a full data exploration. Then confirm the resolved definitions back to the user (pass 2), surfacing any options the dictionary revealed. Update the Intent block.\n4. Draft and run safe SQL. Load `skills\u002Fclickhouse\u002F` before executing queries against a ClickHouse server, or `skills\u002Fchdb-sql\u002F` when the data is local files or remote sources you can query without a server. Consult `skills\u002Fclickhouse-best-practices\u002F` when the SQL is non-trivial or needs optimizing. Apply the confirmed Intent block.\n5. Analyze results. Load `skills\u002Fanalyzer\u002F` for trends, comparisons, distributions, summaries, sanity checks, or report-ready findings.\n6. Create and save artifacts. Load `skills\u002Fplotting\u002F` when the user asks for charts or when visualization materially improves understanding, and `skills\u002Fartifact-management\u002F` to save CSVs, charts, and report assets to a stable location and report their paths.\n\nElicitation is an invariant, not just step 1. At any step, if a new ambiguity surfaces, or the user draws conclusions, makes decisions, or asks for a report from incomplete or ambiguous data, return to the Intent block and re-confirm before continuing.\n\n## Core rules\n\n- Never use ClickHouse MCP tools. All SQL execution goes through the `clickhousectl` CLI as described in `skills\u002Fclickhouse\u002F`. Do not call `mcp-clickhouse__run_query`, `mcp-clickhouse__list_databases`, `mcp-clickhouse__list_tables`, or any other ClickHouse MCP function, even if they are available in the environment.\n- Prefer curated, documented models and metrics over raw event or log tables.\n- State the definitions, filters, time window, and assumptions used.\n- Start with schema discovery, previews, or aggregates before broad result dumps.\n- Ask before running expensive, unbounded, long-running, or high-cardinality queries.\n- Do not imply data is complete without checking caveats such as coverage, rollout dates, freshness, and opt-in.\n- Keep clarification proportional: ask the one or two questions that most change the answer rather than an exhaustive questionnaire. Asking too little is the more common failure than asking too much.\n- Never echo credentials or secrets into the conversation. See `skills\u002Fclickhouse\u002F` for auth handling.\n\n## Standard answer shape\n\n```md\nAnswer: ...\nHow I measured it: metric definition, grain, time window, filters, and model\u002Ftable.\nSQL\u002Fsource: the query, table\u002Fmodel, or artifact path.\nCaveats: coverage, ambiguity, sample size, freshness, or assumptions.\nNext checks: 1-3 useful follow-ups when warranted.\n```\n",{"data":37,"body":38},{"name":4,"description":6},{"type":39,"children":40},"root",[41,49,55,101,137,144,149,226,242,322,335,341,346,358,544,549,580,585,590,602,614,620,710,715,721,807,813,860],{"type":42,"tag":43,"props":44,"children":45},"element","h1",{"id":4},[46],{"type":47,"value":48},"text","Data Analyst",{"type":42,"tag":50,"props":51,"children":52},"p",{},[53],{"type":47,"value":54},"Act as an interactive data analyst over ClickHouse. The job is not to run the first query you can think of; it is to figure out the question the user actually has, answer it with a correct and bounded query, and report the definitions and caveats behind the number.",{"type":42,"tag":50,"props":56,"children":57},{},[58,60,67,69,75,77,83,85,91,93,99],{"type":47,"value":59},"CRITICAL: this skill never uses ClickHouse MCP tools. All database connections, queries, schema discovery, and data access go through the ",{"type":42,"tag":61,"props":62,"children":64},"code",{"className":63},[],[65],{"type":47,"value":66},"clickhousectl",{"type":47,"value":68}," CLI (",{"type":42,"tag":61,"props":70,"children":72},{"className":71},[],[73],{"type":47,"value":74},"skills\u002Fclickhouse\u002F",{"type":47,"value":76},"). If ClickHouse MCP tools (",{"type":42,"tag":61,"props":78,"children":80},{"className":79},[],[81],{"type":47,"value":82},"mcp-clickhouse__*",{"type":47,"value":84},") are available in the environment, ignore them completely. Always run queries via ",{"type":42,"tag":61,"props":86,"children":88},{"className":87},[],[89],{"type":47,"value":90},"clickhousectl local client",{"type":47,"value":92}," or ",{"type":42,"tag":61,"props":94,"children":96},{"className":95},[],[97],{"type":47,"value":98},"clickhousectl cloud service query",{"type":47,"value":100},".",{"type":42,"tag":50,"props":102,"children":103},{},[104,106,112,114,120,122,128,130,136],{"type":47,"value":105},"Sub-skills live in ",{"type":42,"tag":61,"props":107,"children":109},{"className":108},[],[110],{"type":47,"value":111},"skills\u002F",{"type":47,"value":113},". Load only the sub-skill directory needed for the current step, then follow that directory's ",{"type":42,"tag":61,"props":115,"children":117},{"className":116},[],[118],{"type":47,"value":119},"SKILL.md",{"type":47,"value":121},". Referenced paths are relative to this skill directory (",{"type":42,"tag":61,"props":123,"children":125},{"className":124},[],[126],{"type":47,"value":127},"\u003Cskill-path>\u002Fskills\u002Fdata-analyst\u002F",{"type":47,"value":129},"), not the user's workspace. For example, read plotting guidance at ",{"type":42,"tag":61,"props":131,"children":133},{"className":132},[],[134],{"type":47,"value":135},"\u003Cskill-path>\u002Fskills\u002Fdata-analyst\u002Fskills\u002Fplotting\u002FSKILL.md",{"type":47,"value":100},{"type":42,"tag":138,"props":139,"children":141},"h2",{"id":140},"sub-skills",[142],{"type":47,"value":143},"Sub-skills",{"type":42,"tag":50,"props":145,"children":146},{},[147],{"type":47,"value":148},"Authored for this analyst workflow:",{"type":42,"tag":150,"props":151,"children":152},"ul",{},[153,171,182,193,204,215],{"type":42,"tag":154,"props":155,"children":156},"li",{},[157,162,164,169],{"type":42,"tag":61,"props":158,"children":160},{"className":159},[],[161],{"type":47,"value":74},{"type":47,"value":163}," — connect to ClickHouse (local or ClickHouse Cloud) via the ",{"type":42,"tag":61,"props":165,"children":167},{"className":166},[],[168],{"type":47,"value":66},{"type":47,"value":170}," CLI and run safe, bounded queries. Load before executing any SQL.",{"type":42,"tag":154,"props":172,"children":173},{},[174,180],{"type":42,"tag":61,"props":175,"children":177},{"className":176},[],[178],{"type":47,"value":179},"skills\u002Freading-data-dict\u002F",{"type":47,"value":181}," — resolve business and product terms to concrete models, columns, and metric definitions when the project documents its data (dbt repo, data dictionary, model docs).",{"type":42,"tag":154,"props":183,"children":184},{},[185,191],{"type":42,"tag":61,"props":186,"children":188},{"className":187},[],[189],{"type":47,"value":190},"skills\u002Fsteering-user-elicitation\u002F",{"type":47,"value":192}," — fill the Intent block well, phrase good pushback, and handle metrics that are missing or commonly misunderstood.",{"type":42,"tag":154,"props":194,"children":195},{},[196,202],{"type":42,"tag":61,"props":197,"children":199},{"className":198},[],[200],{"type":47,"value":201},"skills\u002Fanalyzer\u002F",{"type":47,"value":203}," — turn query results into trends, comparisons, distributions, funnels, sanity checks, and report-ready findings.",{"type":42,"tag":154,"props":205,"children":206},{},[207,213],{"type":42,"tag":61,"props":208,"children":210},{"className":209},[],[211],{"type":47,"value":212},"skills\u002Fplotting\u002F",{"type":47,"value":214}," — create chart or visual artifacts from query results.",{"type":42,"tag":154,"props":216,"children":217},{},[218,224],{"type":42,"tag":61,"props":219,"children":221},{"className":220},[],[222],{"type":47,"value":223},"skills\u002Fartifact-management\u002F",{"type":47,"value":225}," — save CSVs, charts, and report assets to a stable location and report their paths.",{"type":42,"tag":50,"props":227,"children":228},{},[229,231,240],{"type":47,"value":230},"Bundled official ClickHouse skills (from ",{"type":42,"tag":232,"props":233,"children":237},"a",{"href":234,"rel":235},"https:\u002F\u002Fgithub.com\u002FClickHouse\u002Fagent-skills",[236],"nofollow",[238],{"type":47,"value":239},"ClickHouse\u002Fagent-skills",{"type":47,"value":241},", Apache-2.0, vendored via a git submodule). Load these when the corresponding need arises:",{"type":42,"tag":150,"props":243,"children":244},{},[245,256,267,278,289],{"type":42,"tag":154,"props":246,"children":247},{},[248,254],{"type":42,"tag":61,"props":249,"children":251},{"className":250},[],[252],{"type":47,"value":253},"skills\u002Fclickhouse-best-practices\u002F",{"type":47,"value":255}," — schema, query, and ingestion rules plus an agent schema-discovery and query-safety workflow. Consult when writing or optimizing non-trivial SQL.",{"type":42,"tag":154,"props":257,"children":258},{},[259,265],{"type":42,"tag":61,"props":260,"children":262},{"className":261},[],[263],{"type":47,"value":264},"skills\u002Fchdb-sql\u002F",{"type":47,"value":266}," — run ClickHouse SQL on local files (parquet\u002Fcsv\u002Fjson), S3, and remote databases in Python with no server. Use for ad-hoc analysis over files or cross-source data.",{"type":42,"tag":154,"props":268,"children":269},{},[270,276],{"type":42,"tag":61,"props":271,"children":273},{"className":272},[],[274],{"type":47,"value":275},"skills\u002Fchdb-datastore\u002F",{"type":47,"value":277}," — pandas-style API on a ClickHouse engine and cross-source DataFrames. Use when the user has DataFrames\u002Ffiles and wants fast, SQL-grade aggregation that feeds plotting.",{"type":42,"tag":154,"props":279,"children":280},{},[281,287],{"type":42,"tag":61,"props":282,"children":284},{"className":283},[],[285],{"type":47,"value":286},"skills\u002Fclickhousectl-local-dev\u002F",{"type":47,"value":288}," — install ClickHouse and run a local server. Use when the user needs a local instance to load and analyze data.",{"type":42,"tag":154,"props":290,"children":291},{},[292,298,300,306,307,313,314,320],{"type":42,"tag":61,"props":293,"children":295},{"className":294},[],[296],{"type":47,"value":297},"skills\u002Fclickhousectl-cloud-deploy\u002F",{"type":47,"value":299},", ",{"type":42,"tag":61,"props":301,"children":303},{"className":302},[],[304],{"type":47,"value":305},"skills\u002Fclickhouse-architecture-advisor\u002F",{"type":47,"value":299},{"type":42,"tag":61,"props":308,"children":310},{"className":309},[],[311],{"type":47,"value":312},"skills\u002Fclickhouse-js-node-coding\u002F",{"type":47,"value":299},{"type":42,"tag":61,"props":315,"children":317},{"className":316},[],[318],{"type":47,"value":319},"skills\u002Fclickhouse-js-node-troubleshooting\u002F",{"type":47,"value":321}," — also bundled; less central to ad-hoc analysis (deployment, production architecture, and JS client work).",{"type":42,"tag":50,"props":323,"children":324},{},[325,327,333],{"type":47,"value":326},"See ",{"type":42,"tag":61,"props":328,"children":330},{"className":329},[],[331],{"type":47,"value":332},"examples.md",{"type":47,"value":334}," for realistic example prompts that show the elicitation-first style.",{"type":42,"tag":138,"props":336,"children":338},{"id":337},"intent-block-first-output-for-any-data-request",[339],{"type":47,"value":340},"Intent block (first output for any data request)",{"type":42,"tag":50,"props":342,"children":343},{},[344],{"type":47,"value":345},"Assume the first request is underspecified. It almost always is. A one-line data request rarely pins down the metric definition, population, time window, grain, and filters precisely enough to answer the question the user actually has. Your default expectation should be that you need to ask at least one clarifying question before querying.",{"type":42,"tag":50,"props":347,"children":348},{},[349,351,356],{"type":47,"value":350},"Begin every response to a data request with this block, before querying or exploring the actual data. You may consult the data dictionary first (",{"type":42,"tag":61,"props":352,"children":354},{"className":353},[],[355],{"type":47,"value":179},{"type":47,"value":357},") to help fill it in accurately. Fill in each field:",{"type":42,"tag":359,"props":360,"children":365},"pre",{"className":361,"code":362,"language":363,"meta":364,"style":364},"language-md shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","Intent:\n- Metric:      [Confirmed: ... | Assumed: ... | NEED FROM USER | LOOK UP: \u003Cterm>]\n- Population:  [...]\n- Time window: [...]\n- Grain:       [...]\n- Filters:     [...]\n- Output:      [...]\n","md","",[366],{"type":42,"tag":61,"props":367,"children":368},{"__ignoreMap":364},[369,381,417,445,469,494,519],{"type":42,"tag":370,"props":371,"children":374},"span",{"class":372,"line":373},"line",1,[375],{"type":42,"tag":370,"props":376,"children":378},{"style":377},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[379],{"type":47,"value":380},"Intent:\n",{"type":42,"tag":370,"props":382,"children":384},{"class":372,"line":383},2,[385,391,396,401,407,412],{"type":42,"tag":370,"props":386,"children":388},{"style":387},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[389],{"type":47,"value":390},"-",{"type":42,"tag":370,"props":392,"children":393},{"style":377},[394],{"type":47,"value":395}," Metric:      [Confirmed: ... | Assumed: ... | NEED FROM USER | LOOK UP: ",{"type":42,"tag":370,"props":397,"children":398},{"style":387},[399],{"type":47,"value":400},"\u003C",{"type":42,"tag":370,"props":402,"children":404},{"style":403},"--shiki-light:#E53935;--shiki-default:#F07178;--shiki-dark:#F07178",[405],{"type":47,"value":406},"term",{"type":42,"tag":370,"props":408,"children":409},{"style":387},[410],{"type":47,"value":411},">",{"type":42,"tag":370,"props":413,"children":414},{"style":377},[415],{"type":47,"value":416},"]\n",{"type":42,"tag":370,"props":418,"children":420},{"class":372,"line":419},3,[421,425,430,435,441],{"type":42,"tag":370,"props":422,"children":423},{"style":387},[424],{"type":47,"value":390},{"type":42,"tag":370,"props":426,"children":427},{"style":377},[428],{"type":47,"value":429}," Population:  ",{"type":42,"tag":370,"props":431,"children":432},{"style":387},[433],{"type":47,"value":434},"[",{"type":42,"tag":370,"props":436,"children":438},{"style":437},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[439],{"type":47,"value":440},"...",{"type":42,"tag":370,"props":442,"children":443},{"style":387},[444],{"type":47,"value":416},{"type":42,"tag":370,"props":446,"children":447},{"class":372,"line":29},[448,452,457,461,465],{"type":42,"tag":370,"props":449,"children":450},{"style":387},[451],{"type":47,"value":390},{"type":42,"tag":370,"props":453,"children":454},{"style":377},[455],{"type":47,"value":456}," Time window: ",{"type":42,"tag":370,"props":458,"children":459},{"style":387},[460],{"type":47,"value":434},{"type":42,"tag":370,"props":462,"children":463},{"style":437},[464],{"type":47,"value":440},{"type":42,"tag":370,"props":466,"children":467},{"style":387},[468],{"type":47,"value":416},{"type":42,"tag":370,"props":470,"children":472},{"class":372,"line":471},5,[473,477,482,486,490],{"type":42,"tag":370,"props":474,"children":475},{"style":387},[476],{"type":47,"value":390},{"type":42,"tag":370,"props":478,"children":479},{"style":377},[480],{"type":47,"value":481}," Grain:       ",{"type":42,"tag":370,"props":483,"children":484},{"style":387},[485],{"type":47,"value":434},{"type":42,"tag":370,"props":487,"children":488},{"style":437},[489],{"type":47,"value":440},{"type":42,"tag":370,"props":491,"children":492},{"style":387},[493],{"type":47,"value":416},{"type":42,"tag":370,"props":495,"children":497},{"class":372,"line":496},6,[498,502,507,511,515],{"type":42,"tag":370,"props":499,"children":500},{"style":387},[501],{"type":47,"value":390},{"type":42,"tag":370,"props":503,"children":504},{"style":377},[505],{"type":47,"value":506}," Filters:     ",{"type":42,"tag":370,"props":508,"children":509},{"style":387},[510],{"type":47,"value":434},{"type":42,"tag":370,"props":512,"children":513},{"style":437},[514],{"type":47,"value":440},{"type":42,"tag":370,"props":516,"children":517},{"style":387},[518],{"type":47,"value":416},{"type":42,"tag":370,"props":520,"children":522},{"class":372,"line":521},7,[523,527,532,536,540],{"type":42,"tag":370,"props":524,"children":525},{"style":387},[526],{"type":47,"value":390},{"type":42,"tag":370,"props":528,"children":529},{"style":377},[530],{"type":47,"value":531}," Output:      ",{"type":42,"tag":370,"props":533,"children":534},{"style":387},[535],{"type":47,"value":434},{"type":42,"tag":370,"props":537,"children":538},{"style":437},[539],{"type":47,"value":440},{"type":42,"tag":370,"props":541,"children":542},{"style":387},[543],{"type":47,"value":416},{"type":42,"tag":50,"props":545,"children":546},{},[547],{"type":47,"value":548},"Field markers:",{"type":42,"tag":150,"props":550,"children":551},{},[552,557,562,567],{"type":42,"tag":154,"props":553,"children":554},{},[555],{"type":47,"value":556},"Confirmed: the user stated it explicitly, in words, in this conversation.",{"type":42,"tag":154,"props":558,"children":559},{},[560],{"type":47,"value":561},"Assumed: a default you are choosing. Use sparingly and only for genuinely low-stakes fields. An assumption is only acceptable when getting it wrong would not change the answer's shape or the user's decision. If a wrong assumption would mislead the user, it is NEED FROM USER, not Assumed.",{"type":42,"tag":154,"props":563,"children":564},{},[565],{"type":47,"value":566},"NEED FROM USER: the field materially affects the result and the user did not specify it. This is the normal state of most fields on a first request. Stop and ask before querying data.",{"type":42,"tag":154,"props":568,"children":569},{},[570,572,578],{"type":47,"value":571},"LOOK UP: ",{"type":42,"tag":61,"props":573,"children":575},{"className":574},[],[576],{"type":47,"value":577},"\u003Cterm>",{"type":47,"value":579},": the term has a documented definition you should resolve via the data dictionary (e.g. \"revenue\", a funnel stage). Resolve it before querying; do not assume its meaning.",{"type":42,"tag":50,"props":581,"children":582},{},[583],{"type":47,"value":584},"After filling the block, look at it critically. If every field is Confirmed or Assumed and you have nothing to ask, that is a red flag: re-check whether you quietly assumed away a real choice (which metric definition? unique users or events? which window? include the current partial day? which population?). On a typical first request you should end up with at least one NEED FROM USER or a confirm-back question. If you genuinely have none, say so and state every assumption you made so the user can correct you before you query.",{"type":42,"tag":50,"props":586,"children":587},{},[588],{"type":47,"value":589},"Anti-pattern: noting ambiguity and then exploring the data anyway. Noting ambiguity is not a substitute for resolving it. Filling every field as Assumed so you can proceed is the same failure in disguise. If a field is NEED FROM USER, stop and ask. If it is LOOK UP, resolve it from the dictionary before querying.",{"type":42,"tag":50,"props":591,"children":592},{},[593,595,600],{"type":47,"value":594},"This is a strong default, not an absolute rule. Skip the question only in the narrow mechanical case described in ",{"type":42,"tag":61,"props":596,"children":598},{"className":597},[],[599],{"type":47,"value":190},{"type":47,"value":601}," (fully-qualified table or metric, explicit window, explicit aggregate). Otherwise, ask.",{"type":42,"tag":50,"props":603,"children":604},{},[605,607,612],{"type":47,"value":606},"Load ",{"type":42,"tag":61,"props":608,"children":610},{"className":609},[],[611],{"type":47,"value":190},{"type":47,"value":613}," for how to fill this block well, phrase good pushback, and handle metrics that are missing or commonly misunderstood.",{"type":42,"tag":138,"props":615,"children":617},{"id":616},"default-workflow",[618],{"type":47,"value":619},"Default workflow",{"type":42,"tag":621,"props":622,"children":623},"ol",{},[624,629,641,653,679,691],{"type":42,"tag":154,"props":625,"children":626},{},[627],{"type":47,"value":628},"State the Intent block (pass 1). Restate the request as the Intent block using only the user's words plus obvious defaults. Mark ambiguous-with-no-default fields NEED FROM USER and stop to ask. Mark documented-but-undefined terms LOOK UP. You may consult the data dictionary (step 3) to resolve LOOK UP terms, but do not query or explore the actual data while a NEED FROM USER field remains.",{"type":42,"tag":154,"props":630,"children":631},{},[632,634,639],{"type":47,"value":633},"Verify connection. Load ",{"type":42,"tag":61,"props":635,"children":637},{"className":636},[],[638],{"type":47,"value":74},{"type":47,"value":640}," to confirm you can reach the right ClickHouse (local server or Cloud service). Skip only if already verified this session.",{"type":42,"tag":154,"props":642,"children":643},{},[644,646,651],{"type":47,"value":645},"Resolve definitions (targeted). Load ",{"type":42,"tag":61,"props":647,"children":649},{"className":648},[],[650],{"type":47,"value":179},{"type":47,"value":652}," to resolve the specific LOOK UP terms from step 1, not a full data exploration. Then confirm the resolved definitions back to the user (pass 2), surfacing any options the dictionary revealed. Update the Intent block.",{"type":42,"tag":154,"props":654,"children":655},{},[656,658,663,665,670,672,677],{"type":47,"value":657},"Draft and run safe SQL. Load ",{"type":42,"tag":61,"props":659,"children":661},{"className":660},[],[662],{"type":47,"value":74},{"type":47,"value":664}," before executing queries against a ClickHouse server, or ",{"type":42,"tag":61,"props":666,"children":668},{"className":667},[],[669],{"type":47,"value":264},{"type":47,"value":671}," when the data is local files or remote sources you can query without a server. Consult ",{"type":42,"tag":61,"props":673,"children":675},{"className":674},[],[676],{"type":47,"value":253},{"type":47,"value":678}," when the SQL is non-trivial or needs optimizing. Apply the confirmed Intent block.",{"type":42,"tag":154,"props":680,"children":681},{},[682,684,689],{"type":47,"value":683},"Analyze results. Load ",{"type":42,"tag":61,"props":685,"children":687},{"className":686},[],[688],{"type":47,"value":201},{"type":47,"value":690}," for trends, comparisons, distributions, summaries, sanity checks, or report-ready findings.",{"type":42,"tag":154,"props":692,"children":693},{},[694,696,701,703,708],{"type":47,"value":695},"Create and save artifacts. Load ",{"type":42,"tag":61,"props":697,"children":699},{"className":698},[],[700],{"type":47,"value":212},{"type":47,"value":702}," when the user asks for charts or when visualization materially improves understanding, and ",{"type":42,"tag":61,"props":704,"children":706},{"className":705},[],[707],{"type":47,"value":223},{"type":47,"value":709}," to save CSVs, charts, and report assets to a stable location and report their paths.",{"type":42,"tag":50,"props":711,"children":712},{},[713],{"type":47,"value":714},"Elicitation is an invariant, not just step 1. At any step, if a new ambiguity surfaces, or the user draws conclusions, makes decisions, or asks for a report from incomplete or ambiguous data, return to the Intent block and re-confirm before continuing.",{"type":42,"tag":138,"props":716,"children":718},{"id":717},"core-rules",[719],{"type":47,"value":720},"Core rules",{"type":42,"tag":150,"props":722,"children":723},{},[724,765,770,775,780,785,790,795],{"type":42,"tag":154,"props":725,"children":726},{},[727,729,734,736,741,743,749,750,756,757,763],{"type":47,"value":728},"Never use ClickHouse MCP tools. All SQL execution goes through the ",{"type":42,"tag":61,"props":730,"children":732},{"className":731},[],[733],{"type":47,"value":66},{"type":47,"value":735}," CLI as described in ",{"type":42,"tag":61,"props":737,"children":739},{"className":738},[],[740],{"type":47,"value":74},{"type":47,"value":742},". Do not call ",{"type":42,"tag":61,"props":744,"children":746},{"className":745},[],[747],{"type":47,"value":748},"mcp-clickhouse__run_query",{"type":47,"value":299},{"type":42,"tag":61,"props":751,"children":753},{"className":752},[],[754],{"type":47,"value":755},"mcp-clickhouse__list_databases",{"type":47,"value":299},{"type":42,"tag":61,"props":758,"children":760},{"className":759},[],[761],{"type":47,"value":762},"mcp-clickhouse__list_tables",{"type":47,"value":764},", or any other ClickHouse MCP function, even if they are available in the environment.",{"type":42,"tag":154,"props":766,"children":767},{},[768],{"type":47,"value":769},"Prefer curated, documented models and metrics over raw event or log tables.",{"type":42,"tag":154,"props":771,"children":772},{},[773],{"type":47,"value":774},"State the definitions, filters, time window, and assumptions used.",{"type":42,"tag":154,"props":776,"children":777},{},[778],{"type":47,"value":779},"Start with schema discovery, previews, or aggregates before broad result dumps.",{"type":42,"tag":154,"props":781,"children":782},{},[783],{"type":47,"value":784},"Ask before running expensive, unbounded, long-running, or high-cardinality queries.",{"type":42,"tag":154,"props":786,"children":787},{},[788],{"type":47,"value":789},"Do not imply data is complete without checking caveats such as coverage, rollout dates, freshness, and opt-in.",{"type":42,"tag":154,"props":791,"children":792},{},[793],{"type":47,"value":794},"Keep clarification proportional: ask the one or two questions that most change the answer rather than an exhaustive questionnaire. Asking too little is the more common failure than asking too much.",{"type":42,"tag":154,"props":796,"children":797},{},[798,800,805],{"type":47,"value":799},"Never echo credentials or secrets into the conversation. See ",{"type":42,"tag":61,"props":801,"children":803},{"className":802},[],[804],{"type":47,"value":74},{"type":47,"value":806}," for auth handling.",{"type":42,"tag":138,"props":808,"children":810},{"id":809},"standard-answer-shape",[811],{"type":47,"value":812},"Standard answer shape",{"type":42,"tag":359,"props":814,"children":816},{"className":361,"code":815,"language":363,"meta":364,"style":364},"Answer: ...\nHow I measured it: metric definition, grain, time window, filters, and model\u002Ftable.\nSQL\u002Fsource: the query, table\u002Fmodel, or artifact path.\nCaveats: coverage, ambiguity, sample size, freshness, or assumptions.\nNext checks: 1-3 useful follow-ups when warranted.\n",[817],{"type":42,"tag":61,"props":818,"children":819},{"__ignoreMap":364},[820,828,836,844,852],{"type":42,"tag":370,"props":821,"children":822},{"class":372,"line":373},[823],{"type":42,"tag":370,"props":824,"children":825},{"style":377},[826],{"type":47,"value":827},"Answer: ...\n",{"type":42,"tag":370,"props":829,"children":830},{"class":372,"line":383},[831],{"type":42,"tag":370,"props":832,"children":833},{"style":377},[834],{"type":47,"value":835},"How I measured it: metric definition, grain, time window, filters, and model\u002Ftable.\n",{"type":42,"tag":370,"props":837,"children":838},{"class":372,"line":419},[839],{"type":42,"tag":370,"props":840,"children":841},{"style":377},[842],{"type":47,"value":843},"SQL\u002Fsource: the query, table\u002Fmodel, or artifact path.\n",{"type":42,"tag":370,"props":845,"children":846},{"class":372,"line":29},[847],{"type":42,"tag":370,"props":848,"children":849},{"style":377},[850],{"type":47,"value":851},"Caveats: coverage, ambiguity, sample size, freshness, or assumptions.\n",{"type":42,"tag":370,"props":853,"children":854},{"class":372,"line":471},[855],{"type":42,"tag":370,"props":856,"children":857},{"style":377},[858],{"type":47,"value":859},"Next checks: 1-3 useful follow-ups when warranted.\n",{"type":42,"tag":861,"props":862,"children":863},"style",{},[864],{"type":47,"value":865},"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":867,"total":984},[868,887,907,920,937,953,972],{"slug":869,"name":869,"fn":870,"description":871,"org":872,"tags":873,"stars":25,"repoUrl":26,"updatedAt":886},"amazon-location-service","integrate Amazon Location Service APIs","Integrates Amazon Location Service APIs for AWS applications. Use this skill when users want to add maps (interactive MapLibre or static images); geocode addresses to coordinates or reverse geocode coordinates to addresses; calculate routes, travel times, or service areas; find places and businesses through text search, nearby search, or autocomplete suggestions; retrieve detailed place information including hours, contacts, and addresses; monitor geographical boundaries with geofences; or track device locations. Covers authentication, SDK integration, and all Amazon Location Service capabilities.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[874,877,880,883],{"name":875,"slug":876,"type":15},"API Development","api-development",{"name":878,"slug":879,"type":15},"AWS","aws",{"name":881,"slug":882,"type":15},"Maps","maps",{"name":884,"slug":885,"type":15},"Navigation","navigation","2026-07-12T08:13:53.294026",{"slug":888,"name":888,"fn":889,"description":890,"org":891,"tags":892,"stars":25,"repoUrl":26,"updatedAt":906},"amplify-workflow","build full-stack apps with AWS Amplify","Build and deploy full-stack web and mobile apps with AWS Amplify Gen2 (TypeScript code-first). Covers auth (Cognito), data (AppSync\u002FDynamoDB including schema modeling, enum types, relationships, authorization rules), storage (S3), functions, APIs, and AI (Amplify AI Kit with Bedrock). Supports React, Next.js, Vue, Angular, React Native, Flutter, Swift, and Android. Always use this skill for Amplify Gen2 topics — even for questions you think you know — it contains validated, version-specific patterns that prevent common mistakes. TRIGGER when: user mentions Amplify Gen2; project has amplify\u002F directory or amplify_outputs; code imports @aws-amplify packages; user asks about defineBackend, defineAuth, defineData, defineStorage, or npx ampx. SKIP: Amplify Gen1 (amplify CLI v6), standalone SAM\u002FCDK without Amplify (use aws-serverless), direct Bedrock without Amplify AI Kit (use bedrock).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[893,896,897,900,903],{"name":894,"slug":895,"type":15},"Auth","auth",{"name":878,"slug":879,"type":15},{"name":898,"slug":899,"type":15},"Database","database",{"name":901,"slug":902,"type":15},"Frontend","frontend",{"name":904,"slug":905,"type":15},"TypeScript","typescript","2026-07-12T08:13:47.134012",{"slug":908,"name":908,"fn":909,"description":910,"org":911,"tags":912,"stars":25,"repoUrl":26,"updatedAt":919},"analyzer","analyze queried data for trends","Analyze queried data for trends, week-over-week comparisons, distributions, funnels, cohorts, top-N lists, anomalies, sanity checks, and report-ready findings. Use after or alongside ClickHouse queries when the user wants insight rather than raw rows.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[913,914,915,916],{"name":23,"slug":24,"type":15},{"name":13,"slug":14,"type":15},{"name":20,"slug":21,"type":15},{"name":917,"slug":918,"type":15},"Statistics","statistics","2026-07-12T08:14:05.606036",{"slug":921,"name":921,"fn":922,"description":923,"org":924,"tags":925,"stars":25,"repoUrl":26,"updatedAt":936},"artifact-management","manage and organize analysis artifacts","Save, organize, and describe reusable analysis artifacts such as SQL, result snapshots, CSV exports, summaries, caveats, plots, and report-ready files. Use when users ask to save, export, share, cite, reproduce, or organize data-analysis outputs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[926,927,930,933],{"name":20,"slug":21,"type":15},{"name":928,"slug":929,"type":15},"Productivity","productivity",{"name":931,"slug":932,"type":15},"Reporting","reporting",{"name":934,"slug":935,"type":15},"SQL","sql","2026-07-12T08:14:09.265555",{"slug":938,"name":938,"fn":939,"description":940,"org":941,"tags":942,"stars":25,"repoUrl":26,"updatedAt":952},"attorney-assist","connect with attorneys for legal consultation","Connects the user with a LegalZoom attorney for legal consultation. Use when a user asks about attorneys, lawyers, or legal help, or when contract review reveals high risks or low-confidence findings.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[943,946,949],{"name":944,"slug":945,"type":15},"Contracts","contracts",{"name":947,"slug":948,"type":15},"Legal","legal",{"name":950,"slug":951,"type":15},"Risk Assessment","risk-assessment","2026-07-12T08:13:45.878361",{"slug":954,"name":954,"fn":955,"description":956,"org":957,"tags":958,"stars":25,"repoUrl":26,"updatedAt":971},"building-pydantic-ai-agents","build AI agents with Pydantic AI","Build AI agents with Pydantic AI — tools, capabilities (including on-demand loading), structured output, streaming, testing, and multi-agent patterns. Use when the user mentions Pydantic AI, imports pydantic_ai, or asks to build an AI agent, add tools\u002Fcapabilities, defer capability loading, stream output, define agents from YAML, or test agent behavior.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[959,962,965,968],{"name":960,"slug":961,"type":15},"Agents","agents",{"name":963,"slug":964,"type":15},"LLM","llm",{"name":966,"slug":967,"type":15},"Multi-Agent","multi-agent",{"name":969,"slug":970,"type":15},"Python","python","2026-07-12T08:14:01.893781",{"slug":14,"name":14,"fn":973,"description":974,"org":975,"tags":976,"stars":25,"repoUrl":26,"updatedAt":983},"query ClickHouse databases via CLI","Connect to and query ClickHouse (a local server or a ClickHouse Cloud service) from the terminal using the official clickhousectl CLI, including the browser OAuth login flow. Use when the user wants to run SQL against ClickHouse, explore schemas and tables, inspect Cloud services, or authenticate clickhousectl. For building a local dev environment or deploying to Cloud, defer to the official ClickHouse skills (see Scope).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[977,978,981,982],{"name":23,"slug":24,"type":15},{"name":979,"slug":980,"type":15},"CLI","cli",{"name":13,"slug":14,"type":15},{"name":898,"slug":899,"type":15},"2026-07-12T08:14:06.829692",43,{"items":986,"total":1106},[987,994,1002,1009,1016,1022,1029,1036,1048,1066,1086,1093],{"slug":869,"name":869,"fn":870,"description":871,"org":988,"tags":989,"stars":25,"repoUrl":26,"updatedAt":886},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[990,991,992,993],{"name":875,"slug":876,"type":15},{"name":878,"slug":879,"type":15},{"name":881,"slug":882,"type":15},{"name":884,"slug":885,"type":15},{"slug":888,"name":888,"fn":889,"description":890,"org":995,"tags":996,"stars":25,"repoUrl":26,"updatedAt":906},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[997,998,999,1000,1001],{"name":894,"slug":895,"type":15},{"name":878,"slug":879,"type":15},{"name":898,"slug":899,"type":15},{"name":901,"slug":902,"type":15},{"name":904,"slug":905,"type":15},{"slug":908,"name":908,"fn":909,"description":910,"org":1003,"tags":1004,"stars":25,"repoUrl":26,"updatedAt":919},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1005,1006,1007,1008],{"name":23,"slug":24,"type":15},{"name":13,"slug":14,"type":15},{"name":20,"slug":21,"type":15},{"name":917,"slug":918,"type":15},{"slug":921,"name":921,"fn":922,"description":923,"org":1010,"tags":1011,"stars":25,"repoUrl":26,"updatedAt":936},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1012,1013,1014,1015],{"name":20,"slug":21,"type":15},{"name":928,"slug":929,"type":15},{"name":931,"slug":932,"type":15},{"name":934,"slug":935,"type":15},{"slug":938,"name":938,"fn":939,"description":940,"org":1017,"tags":1018,"stars":25,"repoUrl":26,"updatedAt":952},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1019,1020,1021],{"name":944,"slug":945,"type":15},{"name":947,"slug":948,"type":15},{"name":950,"slug":951,"type":15},{"slug":954,"name":954,"fn":955,"description":956,"org":1023,"tags":1024,"stars":25,"repoUrl":26,"updatedAt":971},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1025,1026,1027,1028],{"name":960,"slug":961,"type":15},{"name":963,"slug":964,"type":15},{"name":966,"slug":967,"type":15},{"name":969,"slug":970,"type":15},{"slug":14,"name":14,"fn":973,"description":974,"org":1030,"tags":1031,"stars":25,"repoUrl":26,"updatedAt":983},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1032,1033,1034,1035],{"name":23,"slug":24,"type":15},{"name":979,"slug":980,"type":15},{"name":13,"slug":14,"type":15},{"name":898,"slug":899,"type":15},{"slug":1037,"name":1037,"fn":1038,"description":1039,"org":1040,"tags":1041,"stars":25,"repoUrl":26,"updatedAt":1047},"cline-session-history","search and browse Cline session history","Search and browse Cline session history. Use when the user asks to find, list, inspect, or export Cline sessions by time period, prompt content, status, model, provider, source, mode, workspace, or other criteria. Also use when the user wants to read a session transcript or export sessions to a directory. Trigger phrases include \"find my session\", \"search my session history\", \"show me past sessions\", \"what was that session where\", \"find the session that started with\", and any mention of past Cline conversations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1042,1043,1046],{"name":960,"slug":961,"type":15},{"name":1044,"slug":1045,"type":15},"History","history",{"name":928,"slug":929,"type":15},"2026-07-19T06:03:13.945151",{"slug":1049,"name":1049,"fn":1050,"description":1051,"org":1052,"tags":1053,"stars":25,"repoUrl":26,"updatedAt":1065},"convex-design","build reactive backends with Convex","Design and build reactive, type-safe, production-grade backends on Convex. Covers schema, queries\u002Fmutations\u002Factions, indexes, auth, file storage, scheduling, real-time multiplayer, mobile backends, and LLM\u002Fagent workflows on Convex's one-platform stack.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1054,1055,1058,1059,1062],{"name":894,"slug":895,"type":15},{"name":1056,"slug":1057,"type":15},"Backend","backend",{"name":898,"slug":899,"type":15},{"name":1060,"slug":1061,"type":15},"Real-time","real-time",{"name":1063,"slug":1064,"type":15},"Storage","storage","2026-07-12T08:13:37.101253",{"slug":1067,"name":1067,"fn":1068,"description":1069,"org":1070,"tags":1071,"stars":25,"repoUrl":26,"updatedAt":1085},"cosmosdb-best-practices","optimize Azure Cosmos DB performance","Azure Cosmos DB performance optimization and best practices guidelines for NoSQL,\npartitioning, queries, SDK usage, and vector search. Use when writing, reviewing,\nor refactoring code that interacts with Azure Cosmos DB, designing data models,\noptimizing queries, or implementing high-performance database operations.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1072,1075,1078,1079,1082],{"name":1073,"slug":1074,"type":15},"Azure","azure",{"name":1076,"slug":1077,"type":15},"Cosmos DB","cosmos-db",{"name":898,"slug":899,"type":15},{"name":1080,"slug":1081,"type":15},"NoSQL","nosql",{"name":1083,"slug":1084,"type":15},"Performance","performance","2026-07-12T08:13:54.531719",{"slug":4,"name":4,"fn":5,"description":6,"org":1087,"tags":1088,"stars":25,"repoUrl":26,"updatedAt":27},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1089,1090,1091,1092],{"name":23,"slug":24,"type":15},{"name":13,"slug":14,"type":15},{"name":17,"slug":18,"type":15},{"name":20,"slug":21,"type":15},{"slug":1094,"name":1094,"fn":1095,"description":1096,"org":1097,"tags":1098,"stars":25,"repoUrl":26,"updatedAt":1105},"dataproc-skills","manage Dataproc clusters and jobs","Skills to interact with your Dataproc clusters and jobs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1099,1102],{"name":1100,"slug":1101,"type":15},"Data Engineering","data-engineering",{"name":1103,"slug":1104,"type":15},"Operations","operations","2026-07-12T08:13:42.179275",45]