[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BYPA6RF58NCPAMASRX4EF4":3,"skill-same:01M7BYPA6RF58NCPAMASRX4EF4":43},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":13,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M7BYPA6RF58NCPAMASRX4EF4","integrate-hub-agent",null,"Embeds one of this repo's pre-built hub agents (under hub\u002Fagents\u002F) into a developer's own application. Use when a developer wants to integrate, add, embed, or wire a GAIA hub agent into their app, project, or codebase — e.g. 'integrate a hub agent into my app', 'add the \u003Cname> agent to my project', 'embed a GAIA agent', 'how do I use the email\u002Fanalyst\u002Fcode agent in my code'. Not for authoring a new agent (use gaia-agent-builder) or releasing\u002Fpublishing one (use agent-hub-release).","secure-workflows\u002Fgaia","secure-workflows","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F311422533?v=4","01M7BYM925H8FAXM1DRD7F5DQD","01M7BYM92V01EZBAV87C93HNKX",0,"MIT","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fintegrate-hub-agent\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: \"integrate-hub-agent\"\ndescription: \"Embeds one of this repo's pre-built hub agents (under hub\u002Fagents\u002F) into a developer's own application. Use when a developer wants to integrate, add, embed, or wire a GAIA hub agent into their app, project, or codebase — e.g. 'integrate a hub agent into my app', 'add the \u003Cname> agent to my project', 'embed a GAIA agent', 'how do I use the email\u002Fanalyst\u002Fcode agent in my code'. Not for authoring a new agent (use gaia-agent-builder) or releasing\u002Fpublishing one (use agent-hub-release).\"\n---\n\n# Integrating a GAIA Hub Agent\n\nThis repo's `hub\u002F` ships pre-built agents, each packaged for a developer to drop into\n**their own** application. The work: learn the layout and the two integration shapes,\npin down **which** agent they mean, then **hand off to that agent's own guidance** —\nwhere the real, current, agent-specific steps live.\n\nThe recipe is **agent-agnostic.** Any agent named below (e.g. npm `agent-email`) is a\nlabeled example of its shape, never the subject. Discover the real agent set live —\nnever trust a hardcoded list.\n\n## Step 1 — Discover the current agents\n\nHub agents are one directory per agent, with a runtime subdir inside:\n`hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002F`. **List the tree now** rather than assuming — the set\nchanges:\n\n```bash\nls hub\u002Fagents\u002F                 # one dir per agent (e.g. analyst, code, email, jira, …)\nls hub\u002Fagents\u002Femail\u002F           # an agent's runtimes (e.g. python, npm; cpp when present)\n```\n\nTreat whatever the listing returns as authoritative. A missing runtime subdir just\nmeans the agent doesn't ship that shape today. `hub\u002Fagents\u002FREADME.md` is a useful\ncatalog of the python set and the example agents (`hello-world`, `word-count`,\n`doc-search`) if the developer is unsure what's available.\n\n## Step 2 — Identify which agent the developer wants\n\nPin down a single `\u003Cname>\u002F\u003Cruntime>` before integrating:\n\n- If the developer named an agent (\"the email agent\", \"analyst\"), match it against the\n  listing. Three names can diverge — watch for it: the friendly name, the **directory**\n  (`\u003Cname>`), and the registry **`id`** in `gaia-agent.yaml` need not match (the\n  `analyst` directory registers as id `data`). The directory is how you find the\n  package; the **`id`** is what you pass when invoking a python agent (Step 3).\n- The same capability can ship in more than one runtime (the email agent is npm\n  `agent-email` **and** python `email`). Confirm the runtime, since it picks the shape:\n  a JS\u002FTS\u002FElectron app wants the npm package; a Python app wants the python package.\n- If they described a capability (\"triage my inbox\", \"analyze a CSV\"), skim candidate\n  agents' `gaia-agent.yaml` `description`\u002F`tags` (python) or `package.json`\n  `description` (npm) to find the match, then confirm with the developer.\n\n## Step 3 — Know the two integration shapes\n\nThey are genuinely different. Use the shape to set expectations, but the agent's own\ndocs (Step 4) are the source of truth for exact calls.\n\n### npm agents (`hub\u002Fagents\u002F\u003Cname>\u002Fnpm\u002F`)\n\nThe package is a **client for a local native sidecar**. Shape:\n\n1. `npm install` the package (it is typically ESM-only — `import`, not `require`).\n2. Fetch + **SHA-256-verify** the platform binary (a build-time step), then **spawn\n   it as a local HTTP sidecar** and wait for health.\n3. Call the package's **typed TypeScript client** methods against the running sidecar.\n4. **Shut the sidecar down** on exit (packages may also auto-reap on process exit).\n\nNo Python, no separate GAIA install — the frozen binary carries the agent. The sidecar\nusually serves **same-origin only**, so an Electron renderer drives it via main-process\nIPC, not a cross-origin fetch. The agent still needs a local model backend (Lemonade\nServer) for inference. Reference example: npm `agent-email`.\n\n### python agents (`hub\u002Fagents\u002F\u003Cname>\u002Fpython\u002F`)\n\nThe package is a **GAIA framework plugin**, not a sidecar. Shape:\n\n1. `pip install gaia-agent-\u003Cid>` (it depends on the published `amd-gaia` wheel).\n2. Installing **auto-registers** the agent into the GAIA registry via the\n   `gaia.agent` entry-point group (declared in `pyproject.toml`) — no hardcoded list,\n   no manual wiring. The registry discovers it on import.\n3. Invoke it through GAIA per the agent's `gaia-agent.yaml` manifest and `README.md`.\n   The manifest declares the `id`, the `models` it expects, and which `interfaces`\n   it exposes (`cli` \u002F `api_server` \u002F `mcp_server` \u002F `pipe`) — drive it through whichever\n   the app uses, plus any required env\u002Fconfig. For **in-process** embedding, the base\n   `Agent` entry point is `process_query(text) -> dict`; construct the agent directly\n   (its package exports the class + a `*Config`) or resolve it by id with the registry —\n   which must be populated first: `r = AgentRegistry(); r.discover(); r.create_agent(\"\u003Cid>\")`\n   (`discover()` scans the `gaia.agent` entry points; without it the registry is empty\n   and `create_agent` raises `ValueError`). Confirm the exact entry against the agent's\n   own source (Step 4) rather than assuming — config knobs vary per agent.\n\nThe npm sidecar\u002Fbinary lifecycle (fetch-binary, spawn, shutdown, typed HTTP client)\n**does not apply here** — don't describe a python agent as a sidecar. Depth varies by\nagent; describe only what the agent's own artifacts support rather than padding it to\nmatch npm's lifecycle.\n\n## Step 4 — Route to the agent's own guidance (do this, don't summarize from memory)\n\nPer-agent integration guidance exists **unevenly**. **Check for a per-agent\n`SKILL.md` and branch** — never assume one is there, and never rely on it being\nauto-discovered as a child skill; you must Read it explicitly.\n\n**Branch A — the agent ships its own `SKILL.md`:**\n\n```bash\nls hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md\n```\n\nIf present, **Read `hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md` and carry out its\ninstructions as the integration steps.** That file is itself a skill addressed to\nyou, the assistant — it is the authoritative, agent-specific recipe, not something to\nmerely mention to the developer. Follow its steps directly, and follow any deeper\nfile it points to (e.g. a `SPEC.md` next to it) when its steps call for that detail.\nThe npm `agent-email` package is the current working example: its `SKILL.md`\n(frontmatter `name: integrate-agent-email`) cascades into a `SPEC.md` for the full\ncontract.\n\n**Branch B — no per-agent `SKILL.md`:**\n\nRead the agent's own artifacts **in this priority order** and **synthesize** the\nintegration steps from them (don't just list the files back to the developer):\n\n1. **`README.md`** — the integrator-facing overview: install command, what it does,\n   any usage snippet.\n2. **Runtime manifest** — `gaia-agent.yaml` (and\u002For `pyproject.toml`) for python,\n   `package.json` for npm: the canonical `id`\u002Fpackage name, declared `models`,\n   exposed `interfaces`, dependencies, and entry points.\n3. **Source entry point** — the package's entry module (python: `gaia_agent_\u003Cid>\u002F__init__.py`\n   `build_registration()` → `agent.py`; npm: the `main`\u002F`exports` entry in\n   `package.json`): the ground truth when README and manifest leave a gap.\n\nSynthesize these into concrete steps shaped per Step 3 (install → register\u002Fconfigure →\ninvoke for python; install → start sidecar → call client → shut down for npm). Most\npython agents are currently in this branch.\n\n## Writing it up\n\nFollow [CLAUDE.md → How You Communicate](..\u002F..\u002F..\u002FCLAUDE.md#how-you-communicate): open with\nwhat the agent does for the developer and the one command that installs it, then layer the\nconfig and contract detail underneath. Don't recite the files you read — hand over steps.\n\n## Checklist\n\n- [ ] Listed `hub\u002Fagents\u002F` live — did not trust any hardcoded agent list.\n- [ ] Pinned a single `\u003Cname>\u002F\u003Cruntime>` and confirmed the runtime with the developer.\n- [ ] Set expectations with the correct shape (npm sidecar vs. python framework plugin).\n- [ ] Checked for `hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md`; if present, **Read and\n      executed it** (and any file it references); if absent, read README → manifest →\n      source and synthesized the steps.\n","# Integrating a GAIA Hub Agent\n\nThis repo's `hub\u002F` ships pre-built agents, each packaged for a developer to drop into\n**their own** application. The work: learn the layout and the two integration shapes,\npin down **which** agent they mean, then **hand off to that agent's own guidance** —\nwhere the real, current, agent-specific steps live.\n\nThe recipe is **agent-agnostic.** Any agent named below (e.g. npm `agent-email`) is a\nlabeled example of its shape, never the subject. Discover the real agent set live —\nnever trust a hardcoded list.\n\n## Step 1 — Discover the current agents\n\nHub agents are one directory per agent, with a runtime subdir inside:\n`hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002F`. **List the tree now** rather than assuming — the set\nchanges:\n\n```bash\nls hub\u002Fagents\u002F                 # one dir per agent (e.g. analyst, code, email, jira, …)\nls hub\u002Fagents\u002Femail\u002F           # an agent's runtimes (e.g. python, npm; cpp when present)\n```\n\nTreat whatever the listing returns as authoritative. A missing runtime subdir just\nmeans the agent doesn't ship that shape today. `hub\u002Fagents\u002FREADME.md` is a useful\ncatalog of the python set and the example agents (`hello-world`, `word-count`,\n`doc-search`) if the developer is unsure what's available.\n\n## Step 2 — Identify which agent the developer wants\n\nPin down a single `\u003Cname>\u002F\u003Cruntime>` before integrating:\n\n- If the developer named an agent (\"the email agent\", \"analyst\"), match it against the\n  listing. Three names can diverge — watch for it: the friendly name, the **directory**\n  (`\u003Cname>`), and the registry **`id`** in `gaia-agent.yaml` need not match (the\n  `analyst` directory registers as id `data`). The directory is how you find the\n  package; the **`id`** is what you pass when invoking a python agent (Step 3).\n- The same capability can ship in more than one runtime (the email agent is npm\n  `agent-email` **and** python `email`). Confirm the runtime, since it picks the shape:\n  a JS\u002FTS\u002FElectron app wants the npm package; a Python app wants the python package.\n- If they described a capability (\"triage my inbox\", \"analyze a CSV\"), skim candidate\n  agents' `gaia-agent.yaml` `description`\u002F`tags` (python) or `package.json`\n  `description` (npm) to find the match, then confirm with the developer.\n\n## Step 3 — Know the two integration shapes\n\nThey are genuinely different. Use the shape to set expectations, but the agent's own\ndocs (Step 4) are the source of truth for exact calls.\n\n### npm agents (`hub\u002Fagents\u002F\u003Cname>\u002Fnpm\u002F`)\n\nThe package is a **client for a local native sidecar**. Shape:\n\n1. `npm install` the package (it is typically ESM-only — `import`, not `require`).\n2. Fetch + **SHA-256-verify** the platform binary (a build-time step), then **spawn\n   it as a local HTTP sidecar** and wait for health.\n3. Call the package's **typed TypeScript client** methods against the running sidecar.\n4. **Shut the sidecar down** on exit (packages may also auto-reap on process exit).\n\nNo Python, no separate GAIA install — the frozen binary carries the agent. The sidecar\nusually serves **same-origin only**, so an Electron renderer drives it via main-process\nIPC, not a cross-origin fetch. The agent still needs a local model backend (Lemonade\nServer) for inference. Reference example: npm `agent-email`.\n\n### python agents (`hub\u002Fagents\u002F\u003Cname>\u002Fpython\u002F`)\n\nThe package is a **GAIA framework plugin**, not a sidecar. Shape:\n\n1. `pip install gaia-agent-\u003Cid>` (it depends on the published `amd-gaia` wheel).\n2. Installing **auto-registers** the agent into the GAIA registry via the\n   `gaia.agent` entry-point group (declared in `pyproject.toml`) — no hardcoded list,\n   no manual wiring. The registry discovers it on import.\n3. Invoke it through GAIA per the agent's `gaia-agent.yaml` manifest and `README.md`.\n   The manifest declares the `id`, the `models` it expects, and which `interfaces`\n   it exposes (`cli` \u002F `api_server` \u002F `mcp_server` \u002F `pipe`) — drive it through whichever\n   the app uses, plus any required env\u002Fconfig. For **in-process** embedding, the base\n   `Agent` entry point is `process_query(text) -> dict`; construct the agent directly\n   (its package exports the class + a `*Config`) or resolve it by id with the registry —\n   which must be populated first: `r = AgentRegistry(); r.discover(); r.create_agent(\"\u003Cid>\")`\n   (`discover()` scans the `gaia.agent` entry points; without it the registry is empty\n   and `create_agent` raises `ValueError`). Confirm the exact entry against the agent's\n   own source (Step 4) rather than assuming — config knobs vary per agent.\n\nThe npm sidecar\u002Fbinary lifecycle (fetch-binary, spawn, shutdown, typed HTTP client)\n**does not apply here** — don't describe a python agent as a sidecar. Depth varies by\nagent; describe only what the agent's own artifacts support rather than padding it to\nmatch npm's lifecycle.\n\n## Step 4 — Route to the agent's own guidance (do this, don't summarize from memory)\n\nPer-agent integration guidance exists **unevenly**. **Check for a per-agent\n`SKILL.md` and branch** — never assume one is there, and never rely on it being\nauto-discovered as a child skill; you must Read it explicitly.\n\n**Branch A — the agent ships its own `SKILL.md`:**\n\n```bash\nls hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md\n```\n\nIf present, **Read `hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md` and carry out its\ninstructions as the integration steps.** That file is itself a skill addressed to\nyou, the assistant — it is the authoritative, agent-specific recipe, not something to\nmerely mention to the developer. Follow its steps directly, and follow any deeper\nfile it points to (e.g. a `SPEC.md` next to it) when its steps call for that detail.\nThe npm `agent-email` package is the current working example: its `SKILL.md`\n(frontmatter `name: integrate-agent-email`) cascades into a `SPEC.md` for the full\ncontract.\n\n**Branch B — no per-agent `SKILL.md`:**\n\nRead the agent's own artifacts **in this priority order** and **synthesize** the\nintegration steps from them (don't just list the files back to the developer):\n\n1. **`README.md`** — the integrator-facing overview: install command, what it does,\n   any usage snippet.\n2. **Runtime manifest** — `gaia-agent.yaml` (and\u002For `pyproject.toml`) for python,\n   `package.json` for npm: the canonical `id`\u002Fpackage name, declared `models`,\n   exposed `interfaces`, dependencies, and entry points.\n3. **Source entry point** — the package's entry module (python: `gaia_agent_\u003Cid>\u002F__init__.py`\n   `build_registration()` → `agent.py`; npm: the `main`\u002F`exports` entry in\n   `package.json`): the ground truth when README and manifest leave a gap.\n\nSynthesize these into concrete steps shaped per Step 3 (install → register\u002Fconfigure →\ninvoke for python; install → start sidecar → call client → shut down for npm). Most\npython agents are currently in this branch.\n\n## Writing it up\n\nFollow [CLAUDE.md → How You Communicate](..\u002F..\u002F..\u002FCLAUDE.md#how-you-communicate): open with\nwhat the agent does for the developer and the one command that installs it, then layer the\nconfig and contract detail underneath. Don't recite the files you read — hand over steps.\n\n## Checklist\n\n- [ ] Listed `hub\u002Fagents\u002F` live — did not trust any hardcoded agent list.\n- [ ] Pinned a single `\u003Cname>\u002F\u003Cruntime>` and confirmed the runtime with the developer.\n- [ ] Set expectations with the correct shape (npm sidecar vs. python framework plugin).\n- [ ] Checked for `hub\u002Fagents\u002F\u003Cname>\u002F\u003Cruntime>\u002FSKILL.md`; if present, **Read and\n      executed it** (and any file it references); if absent, read README → manifest →\n      source and synthesized the steps.\n",{"name":5,"description":7},"main","11039b64e036dca33313350b4cd1d6c2a791eb15",".claude\u002Fskills\u002Fintegrate-hub-agent","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia","https:\u002F\u002Fgithub.com\u002Fsecure-workflows","2026-10-26 10:48:34",{"source":8,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add secure-workflows\u002Fgaia --skill integrate-hub-agent",{"tool":35,"command":36},"bunx","bunx skills add secure-workflows\u002Fgaia --skill integrate-hub-agent",{"tool":38,"command":39},"pnpm","pnpm dlx skills add secure-workflows\u002Fgaia --skill integrate-hub-agent","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fintegrate-hub-agent\u002FSKILL.md",true,[],[44,51,58,65,72,79],{"skillNo":45,"name":46,"nameZh":6,"description":47,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":48,"categories":49,"keywords":50,"publishedAt":18},"01M7BYXGPPQJGVVD0DPS2WK3FS","web-search","Search the web via the Brave Search API. Use when the user asks about current events or external facts.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fweb-search\u002FSKILL.md",[],[],{"skillNo":52,"name":53,"nameZh":6,"description":54,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":55,"categories":56,"keywords":57,"publishedAt":18},"01M7BYXDF2TM5JPMEQKMMHW0VJ","triage-support-ticket","Triage an inbound support ticket end to end. Use when the user pastes a support ticket or asks to triage one.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Ftriage-support-ticket\u002FSKILL.md",[],[],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":18},"01M7BYXAHWNBMJ8Y1Q51EZ53K7","tool-mismatch","Declares a tool its tools.py never registers. Used to prove validation fails loudly with no partial load.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Ftool-mismatch\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":18},"01M7BYX6CZ63AM6HB90PPNXKF7","report-archive","Archive a generated report to the user's reports folder and register it with the team's report index. Use when the user asks to file, archive, or publish a finished report.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Freport-archive\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":18},"01M7BYX4PM4HGNZ0QBZ54BTG1N","local-capability","Writes files on the local workspace. Used to prove Phase 1 refuses local-capability permissions.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Flocal-capability\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":18},"01M7BYX1KVPKHHZ5Z3R3BB0QQA","incident-review","Run an incident postmortem. Stands in for a Claude Code skill library imported read-only from .claude\u002Fskills\u002F.","https:\u002F\u002Fgithub.com\u002Fsecure-workflows\u002Fgaia\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fincident-review\u002FSKILL.md",[],[]]