[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CWN5HATYE6ANKSC5THDTVZ":3,"skill-same:01M6CWN5HATYE6ANKSC5THDTVZ":81},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":22,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":59,"commitSha":60,"skillDirectory":61,"forksCount":62,"repositoryHtmlUrl":63,"repositoryHomepageUrl":64,"ownerHtmlUrl":65,"ownerBlogUrl":6,"ownerDescription":66,"updatedAt":67,"install":68,"resources":80},"01M6CWN5HATYE6ANKSC5THDTVZ","contextdb-autopilot",null,"Automatic ContextDB session lifecycle — init, session events, checkpoints, and continuity artifacts. Use when you need AIOS context persistence without prompt injection. NOT for general task execution — only for ContextDB session management. TRIGGER: contextdb、session persist、checkpoint save.","rexleimo\u002Faios","rexleimo","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11493796?v=4","01M6CWKTCXA8E158HVHMGAD9B5","01M6CWKTGWTT9MMTDYQ4B61CGQ",62,"MIT","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fcontextdb-autopilot\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"contextdb autopilot","contextdb","only","aios","zcode","qoder","project","continuity","general","opencode","prompt injection","session","pi","context","checkpoint","save","lifecycle","injection","init","gemini","2026-11-10 10:12:30",false,"---\nname: contextdb-autopilot\ndescription: \"Automatic ContextDB session lifecycle — init, session events, checkpoints, and continuity artifacts. Use when you need AIOS context persistence without prompt injection. NOT for general task execution — only for ContextDB session management. TRIGGER: contextdb、session persist、checkpoint save.\"\n\ninstallCatalogName: contextdb-autopilot\nclients: [codex, claude, hermes, workbuddy, pi, zcode, qoder]\nscopes: [global, project]\ndefaultInstall:\n  global: true\n  project: false\ntags: [aios, contextdb]\nrepoTargets: [codex, claude, gemini, opencode, hermes, workbuddy, pi, zcode, qoder]\n---\n\n# ContextDB Autopilot\n\nWorking directory: project root. Commands assume repo root; `npm run contextdb` runs from `mcp-server\u002F`.\n\n## Overview\nUse this skill only to persist and retrieve ContextDB state. ContextDB is no longer a prompt-injection layer.\n\nRuntime rules:\n- Do not auto-inject ContextDB packets, continuity summaries, persona overlays, user overlays, route guides, or handoff prompts into model prompts.\n- Interactive startup may show a short local unfinished-task summary to the user; it must not pass that summary as model input.\n- One-shot mode sends only the explicit `--prompt` text to the selected client.\n- Durable workflow constraints belong in checked-in instruction files (`AGENTS.md`, `CLAUDE.md`, `GEMINI.md`) and repo-local skills.\n\nDeprecated prompt-loading flags are intentionally unsupported:\n- `--startup-mode inject`\n- `--context-mode`\n- `--limit`\n- `CTXDB_AUTO_PROMPT`\n\nIf old docs, scripts, or user requests mention these as ways to preload memory, treat them as removed behavior. Use explicit search\u002Ftimeline retrieval instead.\n\nScript path: `scripts\u002Fctx-agent.sh`\n\n## When to Use\n- You want cross-CLI memory continuity (`codex`, `claude`, `gemini`, `opencode`) in the same project.\n- You need automatic session creation, event logging, checkpoints, and continuity\u002Fhandoff artifact writes.\n- You want to retrieve prior facts explicitly before continuing work.\n\nDo not use this skill just to execute a normal coding task. Let the current Rex Capability Command select the task Provider.\n\n## Required Pattern\nUse one-shot mode (`--prompt`) when you want automatic logging around a single explicit request. The prompt is not augmented with ContextDB state.\n\n```bash\nscripts\u002Fctx-agent.sh --agent codex-cli --project rex-cli --prompt \"继续上次任务，先列出你会读取哪些 handoff 文件\"\nscripts\u002Fctx-agent.sh --agent claude-code --project rex-cli --prompt \"记录这个检查点并总结下一步\"\nscripts\u002Fctx-agent.sh --agent gemini-cli --project rex-cli --prompt \"基于我明确给出的文件继续分析\"\nscripts\u002Fctx-agent.sh --agent opencode-cli --project rex-cli --prompt \"执行这个明确请求，不自动加载历史\"\n```\n\n## Session Control\n- Continue same session: `--session \u003Csession_id>`\n- Mark terminal step: `--status done`\n- Disable auto bootstrap for current run: `--no-bootstrap`\n- Disable checkpoint (rare): `--no-checkpoint`\n- Disable auto bootstrap globally: `export AIOS_BOOTSTRAP_AUTO=0`\n- `CODEX_HOME` can be relative; wrappers normalize it against current working directory at runtime.\n\nInteractive wrapper behavior:\n- Direct `codex`\u002F`claude`\u002F`gemini`\u002F`opencode` startup routes through AIOS wrappers without prompt injection.\n- Startup prints only a local unfinished-task summary when tasks\u002Fhandoffs exist.\n- To resume, the user must explicitly name the task\u002Fhandoff to continue before any ContextDB detail is loaded.\n- Route execution defaults to live for team\u002Fsubagent command templates.\n- Override subagent runtime for routed commands with `CTXDB_ROUTE_SUBAGENT_CLIENT=\u003Ccodex-cli|claude-code|gemini-cli|opencode-cli>`.\n\nBootstrap note:\n- On first run in a workspace, `ctx-agent` may auto-create `tasks\u002Fpending\u002Ftask_\u003Ctimestamp>_bootstrap_guidelines\u002F*` and `tasks\u002F.current-task` if both are empty.\n\nExample:\n```bash\nscripts\u002Fctx-agent.sh \\\n  --agent codex-cli \\\n  --project rex-cli \\\n  --session codex-cli-20260303T010101-abcd1234 \\\n  --status done \\\n  --prompt \"所有改动完成，写入检查点并给最终总结\"\n```\n\n## Verification\n- Session files: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002F`\n- Continuity artifact: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002Fcontinuity.json`\n- Handoff artifact: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002Fhandoff.json`\n- Manual report export, when requested: `.aios\u002Fcontext-db\u002Fexports\u002F\u003Csession_id>-context.md`\n\n## Retrieval Workflow\n\nWhen retrieving from ContextDB, run search first and then inspect the timeline only for the matching session. Search finds relevant events by keyword; timeline provides chronological context that explains when and why those events occurred.\n\n1. `npm run contextdb -- search --query \"\u003Cterm>\" --project \u003Cproject>` — find relevant events\n2. `npm run contextdb -- timeline --session \u003Csession_id> --limit 30` — see chronological context\n\nOnly then drill into specific events with `event:get`. If retrieval returns empty, run `index:rebuild` once and retry both commands.\n\nOptional semantic rerank:\n\n```bash\nexport CONTEXTDB_SEMANTIC=1\nexport CONTEXTDB_SEMANTIC_PROVIDER=token\nnpm run contextdb -- search --query \"issue auth\" --project rex-cli --semantic\n```\n\nRecovery:\n- If retrieval returns empty unexpectedly, run `index:rebuild` once.\n- Source-of-truth remains `.aios\u002Fcontext-db\u002Fsessions\u002F*`; sidecar is rebuildable cache.\n\n## Compression contract\n\nWhen a session grows past its budget, compress oldest-first under these rules.\nCompression reshapes what you carry forward; it never deletes session history —\noriginals stay in `.aios\u002Fcontext-db\u002Fsessions\u002F*` for the user and for audit.\n\nTier per event range (you declare the tier, the harness only records it):\n\n- `FULL` — keep verbatim (recent rounds, open blockers, unconsumed handoffs).\n- `PARTIAL` — trim raw tool output, keep conclusions plus evidence refs.\n- `SUMMARY` — one conclusion line plus evidence refs per event cluster.\n- `EXCLUDED` — drop pure noise (heartbeats, duplicates, empty polls).\n\nRules:\n\n- Protect the most recent 2 rounds: never compress them below `PARTIAL`.\n- Summarize, don't truncate: the summary stays agent-visible with evidence\n  refs so the next round can retrieve the original on demand.\n- Every compression writes an auditable event: which range, which tier, and\n  the resulting refs. Compressed content stays recallable through search.\n- State the remaining budget each round; cite it when you choose to compress.\n- A compression event that loses an evidence ref is malformed — re-emit it.\n\n## Manual Report Export\n\nUse `context:pack` only for explicit inspection\u002Fdebug export. Do not feed the report into a model prompt by default. Never paste a full ContextDB Report into a model prompt; if a model needs historical evidence, retrieve only the specific event, checkpoint, or ref the user chose.\n\n```bash\ncd mcp-server\nnpm run contextdb -- context:pack \\\n  --session \u003Csession_id> \\\n  --limit 60 \\\n  --token-budget 1200 \\\n  --token-strategy balanced \\\n  --kinds prompt,response,error \\\n  --refs core.ts,cli.ts\n```\n\nDefaults:\n- Event dedupe in report view is enabled.\n- `--token-strategy` supports `legacy|balanced|aggressive` (recommended: `balanced`).\n- You can disable with `--no-dedupe` when needed for debugging.\n","# ContextDB Autopilot\n\nWorking directory: project root. Commands assume repo root; `npm run contextdb` runs from `mcp-server\u002F`.\n\n## Overview\nUse this skill only to persist and retrieve ContextDB state. ContextDB is no longer a prompt-injection layer.\n\nRuntime rules:\n- Do not auto-inject ContextDB packets, continuity summaries, persona overlays, user overlays, route guides, or handoff prompts into model prompts.\n- Interactive startup may show a short local unfinished-task summary to the user; it must not pass that summary as model input.\n- One-shot mode sends only the explicit `--prompt` text to the selected client.\n- Durable workflow constraints belong in checked-in instruction files (`AGENTS.md`, `CLAUDE.md`, `GEMINI.md`) and repo-local skills.\n\nDeprecated prompt-loading flags are intentionally unsupported:\n- `--startup-mode inject`\n- `--context-mode`\n- `--limit`\n- `CTXDB_AUTO_PROMPT`\n\nIf old docs, scripts, or user requests mention these as ways to preload memory, treat them as removed behavior. Use explicit search\u002Ftimeline retrieval instead.\n\nScript path: `scripts\u002Fctx-agent.sh`\n\n## When to Use\n- You want cross-CLI memory continuity (`codex`, `claude`, `gemini`, `opencode`) in the same project.\n- You need automatic session creation, event logging, checkpoints, and continuity\u002Fhandoff artifact writes.\n- You want to retrieve prior facts explicitly before continuing work.\n\nDo not use this skill just to execute a normal coding task. Let the current Rex Capability Command select the task Provider.\n\n## Required Pattern\nUse one-shot mode (`--prompt`) when you want automatic logging around a single explicit request. The prompt is not augmented with ContextDB state.\n\n```bash\nscripts\u002Fctx-agent.sh --agent codex-cli --project rex-cli --prompt \"继续上次任务，先列出你会读取哪些 handoff 文件\"\nscripts\u002Fctx-agent.sh --agent claude-code --project rex-cli --prompt \"记录这个检查点并总结下一步\"\nscripts\u002Fctx-agent.sh --agent gemini-cli --project rex-cli --prompt \"基于我明确给出的文件继续分析\"\nscripts\u002Fctx-agent.sh --agent opencode-cli --project rex-cli --prompt \"执行这个明确请求，不自动加载历史\"\n```\n\n## Session Control\n- Continue same session: `--session \u003Csession_id>`\n- Mark terminal step: `--status done`\n- Disable auto bootstrap for current run: `--no-bootstrap`\n- Disable checkpoint (rare): `--no-checkpoint`\n- Disable auto bootstrap globally: `export AIOS_BOOTSTRAP_AUTO=0`\n- `CODEX_HOME` can be relative; wrappers normalize it against current working directory at runtime.\n\nInteractive wrapper behavior:\n- Direct `codex`\u002F`claude`\u002F`gemini`\u002F`opencode` startup routes through AIOS wrappers without prompt injection.\n- Startup prints only a local unfinished-task summary when tasks\u002Fhandoffs exist.\n- To resume, the user must explicitly name the task\u002Fhandoff to continue before any ContextDB detail is loaded.\n- Route execution defaults to live for team\u002Fsubagent command templates.\n- Override subagent runtime for routed commands with `CTXDB_ROUTE_SUBAGENT_CLIENT=\u003Ccodex-cli|claude-code|gemini-cli|opencode-cli>`.\n\nBootstrap note:\n- On first run in a workspace, `ctx-agent` may auto-create `tasks\u002Fpending\u002Ftask_\u003Ctimestamp>_bootstrap_guidelines\u002F*` and `tasks\u002F.current-task` if both are empty.\n\nExample:\n```bash\nscripts\u002Fctx-agent.sh \\\n  --agent codex-cli \\\n  --project rex-cli \\\n  --session codex-cli-20260303T010101-abcd1234 \\\n  --status done \\\n  --prompt \"所有改动完成，写入检查点并给最终总结\"\n```\n\n## Verification\n- Session files: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002F`\n- Continuity artifact: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002Fcontinuity.json`\n- Handoff artifact: `.aios\u002Fcontext-db\u002Fsessions\u002F\u003Csession_id>\u002Fhandoff.json`\n- Manual report export, when requested: `.aios\u002Fcontext-db\u002Fexports\u002F\u003Csession_id>-context.md`\n\n## Retrieval Workflow\n\nWhen retrieving from ContextDB, run search first and then inspect the timeline only for the matching session. Search finds relevant events by keyword; timeline provides chronological context that explains when and why those events occurred.\n\n1. `npm run contextdb -- search --query \"\u003Cterm>\" --project \u003Cproject>` — find relevant events\n2. `npm run contextdb -- timeline --session \u003Csession_id> --limit 30` — see chronological context\n\nOnly then drill into specific events with `event:get`. If retrieval returns empty, run `index:rebuild` once and retry both commands.\n\nOptional semantic rerank:\n\n```bash\nexport CONTEXTDB_SEMANTIC=1\nexport CONTEXTDB_SEMANTIC_PROVIDER=token\nnpm run contextdb -- search --query \"issue auth\" --project rex-cli --semantic\n```\n\nRecovery:\n- If retrieval returns empty unexpectedly, run `index:rebuild` once.\n- Source-of-truth remains `.aios\u002Fcontext-db\u002Fsessions\u002F*`; sidecar is rebuildable cache.\n\n## Compression contract\n\nWhen a session grows past its budget, compress oldest-first under these rules.\nCompression reshapes what you carry forward; it never deletes session history —\noriginals stay in `.aios\u002Fcontext-db\u002Fsessions\u002F*` for the user and for audit.\n\nTier per event range (you declare the tier, the harness only records it):\n\n- `FULL` — keep verbatim (recent rounds, open blockers, unconsumed handoffs).\n- `PARTIAL` — trim raw tool output, keep conclusions plus evidence refs.\n- `SUMMARY` — one conclusion line plus evidence refs per event cluster.\n- `EXCLUDED` — drop pure noise (heartbeats, duplicates, empty polls).\n\nRules:\n\n- Protect the most recent 2 rounds: never compress them below `PARTIAL`.\n- Summarize, don't truncate: the summary stays agent-visible with evidence\n  refs so the next round can retrieve the original on demand.\n- Every compression writes an auditable event: which range, which tier, and\n  the resulting refs. Compressed content stays recallable through search.\n- State the remaining budget each round; cite it when you choose to compress.\n- A compression event that loses an evidence ref is malformed — re-emit it.\n\n## Manual Report Export\n\nUse `context:pack` only for explicit inspection\u002Fdebug export. Do not feed the report into a model prompt by default. Never paste a full ContextDB Report into a model prompt; if a model needs historical evidence, retrieve only the specific event, checkpoint, or ref the user chose.\n\n```bash\ncd mcp-server\nnpm run contextdb -- context:pack \\\n  --session \u003Csession_id> \\\n  --limit 60 \\\n  --token-budget 1200 \\\n  --token-strategy balanced \\\n  --kinds prompt,response,error \\\n  --refs core.ts,cli.ts\n```\n\nDefaults:\n- Event dedupe in report view is enabled.\n- `--token-strategy` supports `legacy|balanced|aggressive` (recommended: `balanced`).\n- You can disable with `--no-dedupe` when needed for debugging.\n",{"name":5,"tags":48,"scopes":49,"clients":51,"description":7,"repoTargets":56,"defaultInstall":57,"installCatalogName":5},[26,24],[50,29],"global",[52,53,54,55,35,27,28],"codex","claude","hermes","workbuddy",[52,53,42,32,54,55,35,27,28],{"global":58,"project":44},true,"main","125f0c919bcb02c316fd4dcbf10e555fad7ac852","skill-sources\u002Fcontextdb-autopilot",11,"https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios","https:\u002F\u002Fcli.rexai.top\u002F","https:\u002F\u002Fgithub.com\u002Frexleimo","Quickly build your laravel environment","2026-11-07 23:34:17",{"source":8,"skillName":5,"commands":69,"prompt":79,"zipAvailable":58},[70,73,76],{"tool":71,"command":72},"npx","npx skills add rexleimo\u002Faios --skill contextdb-autopilot",{"tool":74,"command":75},"bunx","bunx skills add rexleimo\u002Faios --skill contextdb-autopilot",{"tool":77,"command":78},"pnpm","pnpm dlx skills add rexleimo\u002Faios --skill contextdb-autopilot","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fcontextdb-autopilot\u002FSKILL.md",[],[82,102,120,144,166,186],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":43},"01M6CWQB49G29K3K6DEVQDA61G","versioning-by-impact","Use when completing a task and deciding whether repository changes require a semantic version bump and changelog entry before commit\u002Fpush.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fversioning-by-impact\u002FSKILL.md",[],[89,90,42,91,92,93,94,95,96,35,97,98,32,99,31,100,101,29,28,27],"versioning","versioning by impact","commit","release","changelog","push","repository","changelog entry","use","version","bump","version bump","task",{"skillNo":103,"name":104,"nameZh":6,"description":105,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":106,"categories":107,"keywords":109,"publishedAt":43},"01M6CWQ8HA0ZDKAKS0VFXJZ6TW","verification-loop","Evidence-before-assertions workflow. Use before claiming work is done, before release, and after any behavior change in scripts\u002Fskills\u002FMCP.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fverification-loop\u002FSKILL.md",[108],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[110,111,112,113,42,52,114,115,92,116,117,35,97,118,32,31,119,29,28,27],"verification","loop","verification loop","MCP","workflow","change","work","behavior","evidence","done",{"skillNo":121,"name":122,"nameZh":6,"description":123,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":124,"categories":125,"keywords":132,"publishedAt":43},"01M6CWQ5A61P9QTRK4HRV76MT3","typesafe-ai","Build AI-powered software with TypeSafe: small units of AI intelligence you can use like programming primitives. Its System One models, including Jev, turn natural language and application state into typed judgments and probabilities that code can combine. Use when a feature needs programmable common sense, when brainstorming what AI could make possible in an app, or when an LLM prompt-and-parse step could become a structured decision. Applications include routing, ranking, extraction, verification, and interactive experiences; these are starting points, not the limits. Read live docs and cookbooks to find useful patterns and discover new combinations.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Ftypesafe-ai\u002FSKILL.md",[126,127],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},{"categoryNo":128,"categoryName":129,"taxonomyType":130,"taxonomyTypeDesc":131},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[133,42,134,110,135,136,137,138,35,139,32,140,141,142,29,143,28,27,26,122],"typesafe ai","code","system","decision","common","read","primitives","patterns","natural language","intelligence","discover",{"skillNo":145,"name":146,"nameZh":6,"description":147,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":148,"categories":149,"keywords":151,"publishedAt":43},"01M6CWQ1HAVFV6KQ0CR549Y9AP","skill-opt-lite","Train and optimize skill documents using the SkillOpt methodology — the agent acts as both target and optimizer, running tasks, reflecting on failures, proposing edits, and validating improvements in a self-contained training loop. TRIGGER: optimize skill, train skill, improve skill, skill training, skill optimization, SkillOpt, skill evolution, skill iteration loop, skill feedback loop, skill tuning.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fskill-opt-lite\u002FSKILL.md",[150],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[152,42,153,154,155,156,157,158,159,111,160,32,161,162,163,29,28,27,164,165],"skill opt lite","documents","training","tasks","improve skill","Optimization","tuning","improve","feedback","optimize","self","optimizer","SkillOpt","iteration",{"skillNo":167,"name":168,"nameZh":6,"description":169,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":170,"categories":171,"keywords":173,"publishedAt":43},"01M6CWPX0C8XBBD225VBJRJWSA","skill-constraints","Enforce browser automation safety constraints — tool usage order, anti-detection rules, screenshot discipline, and human-in-the-loop checks. Use when setting up or verifying browser\u002FMCP tool operations, NOT for every skill execution.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fskill-constraints\u002FSKILL.md",[172],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[174,175,113,176,42,177,178,179,180,111,181,35,182,183,32,184,185,29,28,27],"constraints","skill constraints","browser automation","browser","screenshot","Safety","human-in-the-loop","tool","discipline","up","enforce","human",{"skillNo":187,"name":188,"nameZh":6,"description":189,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":190,"categories":191,"keywords":199,"publishedAt":43},"01M6CWPS77PTTK17X8PNJE9RVF","seo-geo-page-optimization","Use when improving a single page for SEO and AI answer-engine visibility (GEO), especially when you need faster first-screen conversion and clearer traffic funnels.","https:\u002F\u002Fgithub.com\u002Frexleimo\u002Faios\u002Fblob\u002Fmain\u002Fskill-sources\u002Fseo-geo-page-optimization\u002FSKILL.md",[192,195,196],{"categoryNo":193,"categoryName":194,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000019","营销推广",{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},{"categoryNo":197,"categoryName":198,"taxonomyType":130,"taxonomyTypeDesc":131},"01JBSEEDCAT000000000000035","市场营销",[157,200,201,202,203,204,42,52,205,206,207,54,50,35,97,32,208,29,28,27],"seo","page","seo geo","geo","seo geo page optimization","content","first","engine","conversion"]