[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AW41KYQCS5G1VRQBSN7TSC":3,"skill-same:01M7AW41KYQCS5G1VRQBSN7TSC":47},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":46},"01M7AW41KYQCS5G1VRQBSN7TSC","chromeapi",null,"Capture a site's internal API routes from your REAL logged-in Chrome and replay them browserless (kuri\u002Funbrowse-style, all local). The deeper-level API layer on top of fcdp. Invoke as \u002Fchromeapi when the user wants to record the XHR\u002Ffetch calls a page makes and re-fire them as plain HTTP with their real cookies — no browser, no telemetry. \u002Fchrome (fcdp) remains the tool for ordinary page driving.","barkleesanders\u002Fclaude-code-starter","barkleesanders","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F30680901?v=4","01M7AW019CGBYWDA8MFJEN1JXC","Barklee Sanders","01M7AW01A5H994ASJ2QZPW67AV",4,"MIT","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fchromeapi\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: chromeapi\nuser-invocable: true\ndescription: Capture a site's internal API routes from your REAL logged-in Chrome and replay them browserless (kuri\u002Funbrowse-style, all local). The deeper-level API layer on top of fcdp. Invoke as \u002Fchromeapi when the user wants to record the XHR\u002Ffetch calls a page makes and re-fire them as plain HTTP with their real cookies — no browser, no telemetry. \u002Fchrome (fcdp) remains the tool for ordinary page driving.\n---\n\n# \u002Fchromeapi — capture & replay a site's internal API (on your real Chrome)\n\nThe **deeper-level API layer** built on top of `\u002Fchrome` (fcdp). Same real logged-in\nDefault-profile Chrome — but instead of clicking\u002Freading a page, it **records the internal\nAPI calls the page fires** into a manifest, then **replays them browserless** as plain `curl`\nusing your real Chrome cookies. This is the local, no-telemetry equivalent of unbrowse\u002Fkuri's\ncapture→replay trick (reverse-engineered 2026-07-22; see memory `reference_fcdp_full_cdp.md`).\n\n`\u002Fchrome` (fcdp) stays exactly as-is for ordinary driving (open\u002Fread\u002Fclick\u002Ftype\u002Fjs\u002Fpdf\u002F…).\n`\u002Fchromeapi` is when the goal is **\"find the API call and replay it without a browser.\"**\n\n## Step 0 — does the site PUBLISH its API? Check before you capture anything\n\n**Run this first, every time.** A capture window is a guess: you get only the endpoints the\npage happened to fire in those N seconds, method+URL only, no schema. If the site publishes\na spec, that's the **entire API surface with parameter names, types and response schemas** —\nstrictly better data, in one request, with no browser.\n\nNote this is the API analog of `\u002Fchrome`'s sitemap-first rule, **not sitemap.xml itself**.\nA sitemap enumerates *pages*; it tells you nothing about an API. The right artifacts here\nare OpenAPI\u002FSwagger\u002FGraphQL. (Do still glance at `robots.txt` — it sometimes names an\n`\u002Fapi-docs` path.)\n\n```bash\n# Prefer the a logged-in browser CLI probe (robots\u002Fsitemap\u002FOpenAPI\u002Fheaders\u002Fscripts):\na logged-in browser CLI inspect https:\u002F\u002F\u003Chost>\n# Old unbrowse MCP spec-discover is UNINSTALLED — do not invoke it.\n# Unpublished \u002F cookie-gated: a logged-in browser CLI capture \u003Curl> [--fcdp]\n#   or ~\u002Ftools\u002Ffcdp-api\u002Ffapi capture \u003Curl> [secs]\n# HAR distill: a logged-in browser CLI distill \u003Cfile.har>\n```\n\nDecision rule:\n- **Spec found** → read it. You now have every route, its params and its response shape.\n  Skip `fapi capture` entirely and go straight to `curl` (still with real cookies for\n  authenticated routes). This is also the only way to learn about endpoints the UI never calls.\n- **GraphQL** (`\u002Fgraphql` responds) → introspection may be enabled; that's the whole schema.\n  Introspection is often disabled in production — if so, fall through to capture.\n- **Nothing published** (the common case for consumer web apps) → proceed to `fapi capture`\n  below. That's the normal path, not a failure.\n\nA 404 on all of the above is a 5-second answer, so there's no reason to skip it — and when\nit hits, it saves the capture cycle *and* gives you schemas `fapi` structurally cannot infer\n(see the \"Gap vs unbrowse\u002Fkuri\" note below: `fcdp intercept` surfaces method+URL, not bodies).\n\n## The tool: `fapi`\n\n`~\u002Ftools\u002Ffcdp-api\u002Ffapi` (also `~\u002Ftools\u002Ffcdp\u002Ffapi`). Built on `fcdp intercept` (capture) +\n`~\u002Ftools\u002Fcookies-txt` (real-profile cookies) + `curl` (browserless replay). Nothing leaves the machine.\n\n```bash\nfapi capture \u003Curl> [secs]   # open a bg tab in REAL Chrome, intercept the API calls it fires -> manifest\nfapi list                   # list captured domains\nfapi show \u003Cdomain>          # print the endpoint manifest (JSON)\nfapi replay \u003Cdomain> [n]    # browserless curl replay of GET endpoint(s), with real Chrome cookies\n```\n\nManifests live in `~\u002F.fcdp-api\u002F\u003Cdomain>.json`: `{domain, trigger, captured_at, endpoints:[{method,url,idempotency}]}`.\n\n## How to run it\n\n1. **Capture** — pick a URL that fires the API you want (the page, or the API URL itself). Give it\n   enough seconds to let the page's XHRs fire (default 8):\n   ```bash\n   ~\u002Ftools\u002Ffcdp-api\u002Ffapi capture \"https:\u002F\u002Fapp.example.com\u002Fdashboard\" 8\n   ```\n   Report the endpoints it found (`fapi show \u003Cdomain>`). If it found 0, the page may fire its\n   XHRs later than the window — re-run with more seconds, or point `capture` directly at the API URL.\n2. **Replay** — re-fire the safe (GET) endpoints without a browser, using the user's real cookies:\n   ```bash\n   ~\u002Ftools\u002Ffcdp-api\u002Ffapi replay app.example.com\n   ```\n   This is step 3 of the web-lookup ladder (cached endpoint replay, no live browser) done against\n   the user's OWN session.\n\n## Rules & safety\n\n- **GET-only auto-replay.** `fapi replay` fires only `GET` endpoints. POST\u002FPUT\u002FDELETE\u002FPATCH are\n  listed but NEVER auto-fired — a mutation is an outward, irreversible action: show the exact\n  request and get explicit per-action approval, then fire it by hand with `curl`.\n- **Real cookies, local only.** `~\u002Ftools\u002Fcookies-txt` reads the Default profile's cookie SQLite +\n  Keychain. Cookies never leave the machine. Only capture\u002Freplay sites the user is lawfully using\n  their own account on.\n- **One debugger per tab.** `fapi capture` uses `fcdp intercept`, which attaches `chrome.debugger`\n  to a fresh background tab and closes it after — it won't fight a tab you're driving with `\u002Fchrome`.\n- **Gap vs unbrowse\u002Fkuri (by design):** stores method+URL only — no response-schema inference or\n  ranked semantic resolve, because `fcdp intercept` surfaces method+URL, not response bodies. If\n  schema inference is ever wanted, add `Network.getResponseBody` capture to fcdp. Better than kuri\n  in the way that matters here: it runs against the user's REAL logins (no separate auth step).\n\n## When to use which\n\n| Want to… | Use |\n|---|---|\n| inspect \u002F fetch \u002F capture \u002F GET-replay \u002F drive as one stack | a logged-in browser CLI (preferred for new work) |\n| open\u002Fread\u002Fclick\u002Ftype\u002Fscreenshot\u002FPDF a page, run JS | `\u002Fchrome` (fcdp) |\n| record a site's internal API calls and replay them browserless | `a logged-in browser CLI capture` + `replay`, or this skill's `fapi` |\n| decompile a binary \u002F RE how some app talks to its API | `\u002Fdecompile` |\n| turn a captured HAR\u002FJSON into a **repeatable assert** (jsonpath on unwrap, status, body keys) | `hurl --test file.hurl` (installed 2026-08-14). Capture with `fhar`\u002F`fapi`\u002Fa logged-in browser CLI, flatten with `gron` to find the path, then write the Hurl file. Never commit the raw HAR. |\n\nFull RE writeup of how this mirrors unbrowse\u002Fkuri: memory `reference_fcdp_full_cdp.md`.\n","# \u002Fchromeapi — capture & replay a site's internal API (on your real Chrome)\n\nThe **deeper-level API layer** built on top of `\u002Fchrome` (fcdp). Same real logged-in\nDefault-profile Chrome — but instead of clicking\u002Freading a page, it **records the internal\nAPI calls the page fires** into a manifest, then **replays them browserless** as plain `curl`\nusing your real Chrome cookies. This is the local, no-telemetry equivalent of unbrowse\u002Fkuri's\ncapture→replay trick (reverse-engineered 2026-07-22; see memory `reference_fcdp_full_cdp.md`).\n\n`\u002Fchrome` (fcdp) stays exactly as-is for ordinary driving (open\u002Fread\u002Fclick\u002Ftype\u002Fjs\u002Fpdf\u002F…).\n`\u002Fchromeapi` is when the goal is **\"find the API call and replay it without a browser.\"**\n\n## Step 0 — does the site PUBLISH its API? Check before you capture anything\n\n**Run this first, every time.** A capture window is a guess: you get only the endpoints the\npage happened to fire in those N seconds, method+URL only, no schema. If the site publishes\na spec, that's the **entire API surface with parameter names, types and response schemas** —\nstrictly better data, in one request, with no browser.\n\nNote this is the API analog of `\u002Fchrome`'s sitemap-first rule, **not sitemap.xml itself**.\nA sitemap enumerates *pages*; it tells you nothing about an API. The right artifacts here\nare OpenAPI\u002FSwagger\u002FGraphQL. (Do still glance at `robots.txt` — it sometimes names an\n`\u002Fapi-docs` path.)\n\n```bash\n# Prefer the a logged-in browser CLI probe (robots\u002Fsitemap\u002FOpenAPI\u002Fheaders\u002Fscripts):\na logged-in browser CLI inspect https:\u002F\u002F\u003Chost>\n# Old unbrowse MCP spec-discover is UNINSTALLED — do not invoke it.\n# Unpublished \u002F cookie-gated: a logged-in browser CLI capture \u003Curl> [--fcdp]\n#   or ~\u002Ftools\u002Ffcdp-api\u002Ffapi capture \u003Curl> [secs]\n# HAR distill: a logged-in browser CLI distill \u003Cfile.har>\n```\n\nDecision rule:\n- **Spec found** → read it. You now have every route, its params and its response shape.\n  Skip `fapi capture` entirely and go straight to `curl` (still with real cookies for\n  authenticated routes). This is also the only way to learn about endpoints the UI never calls.\n- **GraphQL** (`\u002Fgraphql` responds) → introspection may be enabled; that's the whole schema.\n  Introspection is often disabled in production — if so, fall through to capture.\n- **Nothing published** (the common case for consumer web apps) → proceed to `fapi capture`\n  below. That's the normal path, not a failure.\n\nA 404 on all of the above is a 5-second answer, so there's no reason to skip it — and when\nit hits, it saves the capture cycle *and* gives you schemas `fapi` structurally cannot infer\n(see the \"Gap vs unbrowse\u002Fkuri\" note below: `fcdp intercept` surfaces method+URL, not bodies).\n\n## The tool: `fapi`\n\n`~\u002Ftools\u002Ffcdp-api\u002Ffapi` (also `~\u002Ftools\u002Ffcdp\u002Ffapi`). Built on `fcdp intercept` (capture) +\n`~\u002Ftools\u002Fcookies-txt` (real-profile cookies) + `curl` (browserless replay). Nothing leaves the machine.\n\n```bash\nfapi capture \u003Curl> [secs]   # open a bg tab in REAL Chrome, intercept the API calls it fires -> manifest\nfapi list                   # list captured domains\nfapi show \u003Cdomain>          # print the endpoint manifest (JSON)\nfapi replay \u003Cdomain> [n]    # browserless curl replay of GET endpoint(s), with real Chrome cookies\n```\n\nManifests live in `~\u002F.fcdp-api\u002F\u003Cdomain>.json`: `{domain, trigger, captured_at, endpoints:[{method,url,idempotency}]}`.\n\n## How to run it\n\n1. **Capture** — pick a URL that fires the API you want (the page, or the API URL itself). Give it\n   enough seconds to let the page's XHRs fire (default 8):\n   ```bash\n   ~\u002Ftools\u002Ffcdp-api\u002Ffapi capture \"https:\u002F\u002Fapp.example.com\u002Fdashboard\" 8\n   ```\n   Report the endpoints it found (`fapi show \u003Cdomain>`). If it found 0, the page may fire its\n   XHRs later than the window — re-run with more seconds, or point `capture` directly at the API URL.\n2. **Replay** — re-fire the safe (GET) endpoints without a browser, using the user's real cookies:\n   ```bash\n   ~\u002Ftools\u002Ffcdp-api\u002Ffapi replay app.example.com\n   ```\n   This is step 3 of the web-lookup ladder (cached endpoint replay, no live browser) done against\n   the user's OWN session.\n\n## Rules & safety\n\n- **GET-only auto-replay.** `fapi replay` fires only `GET` endpoints. POST\u002FPUT\u002FDELETE\u002FPATCH are\n  listed but NEVER auto-fired — a mutation is an outward, irreversible action: show the exact\n  request and get explicit per-action approval, then fire it by hand with `curl`.\n- **Real cookies, local only.** `~\u002Ftools\u002Fcookies-txt` reads the Default profile's cookie SQLite +\n  Keychain. Cookies never leave the machine. Only capture\u002Freplay sites the user is lawfully using\n  their own account on.\n- **One debugger per tab.** `fapi capture` uses `fcdp intercept`, which attaches `chrome.debugger`\n  to a fresh background tab and closes it after — it won't fight a tab you're driving with `\u002Fchrome`.\n- **Gap vs unbrowse\u002Fkuri (by design):** stores method+URL only — no response-schema inference or\n  ranked semantic resolve, because `fcdp intercept` surfaces method+URL, not response bodies. If\n  schema inference is ever wanted, add `Network.getResponseBody` capture to fcdp. Better than kuri\n  in the way that matters here: it runs against the user's REAL logins (no separate auth step).\n\n## When to use which\n\n| Want to… | Use |\n|---|---|\n| inspect \u002F fetch \u002F capture \u002F GET-replay \u002F drive as one stack | a logged-in browser CLI (preferred for new work) |\n| open\u002Fread\u002Fclick\u002Ftype\u002Fscreenshot\u002FPDF a page, run JS | `\u002Fchrome` (fcdp) |\n| record a site's internal API calls and replay them browserless | `a logged-in browser CLI capture` + `replay`, or this skill's `fapi` |\n| decompile a binary \u002F RE how some app talks to its API | `\u002Fdecompile` |\n| turn a captured HAR\u002FJSON into a **repeatable assert** (jsonpath on unwrap, status, body keys) | `hurl --test file.hurl` (installed 2026-08-14). Capture with `fhar`\u002F`fapi`\u002Fa logged-in browser CLI, flatten with `gron` to find the path, then write the Hurl file. Never commit the raw HAR. |\n\nFull RE writeup of how this mirrors unbrowse\u002Fkuri: memory `reference_fcdp_full_cdp.md`.\n",{"name":5,"description":7,"user-invocable":24},true,"main","e7b39d448c7d905feed2b044d3ccdf60822596ab","skills\u002Fchromeapi",0,"https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter","https:\u002F\u002Fgithub.com\u002Fbarkleesanders","barkleesanders.com","Stay Hungry, Stay Foolish","2026-11-06 08:47:06",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":24},[36,39,42],{"tool":37,"command":38},"npx","npx skills add barkleesanders\u002Fclaude-code-starter --skill chromeapi",{"tool":40,"command":41},"bunx","bunx skills add barkleesanders\u002Fclaude-code-starter --skill chromeapi",{"tool":43,"command":44},"pnpm","pnpm dlx skills add barkleesanders\u002Fclaude-code-starter --skill chromeapi","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fchromeapi\u002FSKILL.md",[],[48,54,61,68,75,82],{"skillNo":49,"name":50,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7AWV75B37B6YKG7ZKGZ783F","zi-search","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fzi-search\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7AWTZA3JGK1T7V1HS543FG7","visualise","Render inline interactive visuals — SVG diagrams, HTML widgets, charts, and explainers — directly in the conversation. Use this skill whenever the user asks to visualize, diagram, chart, illustrate, or explain something visually, or when an explanation would genuinely benefit from a spatial\u002Finteractive diagram rather than text. Also triggers for: flowcharts, architecture diagrams, data visualizations, interactive explainers, comparison layouts, UI mockups, and any request containing 'show me', 'draw', 'map out', 'visualize', or 'diagram'. Even when the user doesn't explicitly ask for a visual, use this skill proactively when the topic has spatial, sequential, or systemic relationships that a diagram would clarify better than prose.","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fvisualise\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7AWTR056C3P9R3VS8YVM2WX","video","When the user wants to create, generate, or produce video content using AI tools or programmatic frameworks. Also use when the user mentions 'video production,' 'AI video,' 'Remotion,' 'Hyperframes,' 'HeyGen,' 'Synthesia,' 'Veo,' 'Runway,' 'Kling,' 'Pika,' 'video generation,' 'AI avatar,' 'talking head video,' 'programmatic video,' 'video template,' 'explainer video,' 'product demo video,' 'video pipeline,' or 'make me a video.' Use this for video creation, generation, and production workflows. For video content strategy and what to post, see social-content. For paid video ad creative, see ad-creative.","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fvideo\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M7AWTHTWBMTYZPMHHEYX3TTG","manim-video","Production pipeline for mathematical and technical animations using Manim Community Edition. Creates 3Blue1Brown-style explainer videos, algorithm visualizations, equation derivations, architecture diagrams, and data stories. Use when users request: animated explanations, math animations, concept visualizations, algorithm walkthroughs, technical explainers, 3Blue1Brown style videos, or any programmatic animation with geometric\u002Fmathematical content.","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fvideo-use\u002Fskills\u002Fmanim-video\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7AWTCXDJ9NWFQW51ENM8A2Z","video-use","Edit any video by conversation. Transcribe, cut, color grade, generate overlay animations, burn subtitles — for talking heads, montages, tutorials, travel, interviews. No presets, no menus. Ask questions, confirm the plan, execute, iterate, persist. Production-correctness rules are hard; everything else is artistic freedom.","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fvideo-use\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":19},"01M7AWTA1D5FP895S33GHSA2G5","ui-improver","https:\u002F\u002Fgithub.com\u002Fbarkleesanders\u002Fclaude-code-starter\u002Fblob\u002Fmain\u002Fskills\u002Fui-improver\u002FSKILL.md",[],[]]