[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M75CE7TPAPCVEHR1X3NW0GF2":3,"skill-same:01M75CE7TPAPCVEHR1X3NW0GF2":48},{"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":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M75CE7TPAPCVEHR1X3NW0GF2","fleet-manager",null,"Use when managing, triaging, restarting, escalating, or summarizing Ghosty Agent Fleet runs and workers.","blissito\u002Fghostycode","blissito","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F156516539?v=4","01M75CD3QJEZD682BR625QD3QA","01M75CD3RB12C1P572X9DRE503",8,"MIT","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fcrates\u002Ftui\u002Fassets\u002Fskills\u002Ffleet-manager\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: fleet-manager\ndescription: Use when managing, triaging, restarting, escalating, or summarizing Ghosty Agent Fleet runs and workers.\nmetadata:\n  short-description: Triage Ghosty Agent Fleet runs\n---\n\n# Fleet Manager\n\nUse this skill when acting as a manager agent for Ghosty Agent Fleet runs.\nYour job is to classify worker state, choose the narrowest safe typed action,\nand leave a ledgered receipt or a safe escalation draft.\n\n## Authority Boundary\n\n- Prefer typed fleet surfaces over shell spelunking: `ghosty fleet status`,\n  `inspect`, `logs`, `artifacts`, `interrupt`, `restart`, `stop`, and the\n  Runtime API fleet endpoints.\n- Do not read `.ghosty\u002Ffleet.jsonl`, host logs, or remote files directly\n  unless the typed command or API is missing required evidence.\n- Do not send Slack, webhook, PagerDuty, email, or chat messages unless the\n  user or run config explicitly authorizes sending. Draft the message instead.\n- Never include secrets, tokens, webhook URLs, routing keys, full prompts, or\n  oversized logs in a summary or escalation.\n\n## Triage Loop\n\n1. Identify the run and worker from the user request, run receipt, or fleet\n   status output. If no worker is named, start with `ghosty fleet status`.\n2. Inspect the worker with `ghosty fleet inspect \u003Cworker-id>` or the matching\n   Runtime API worker endpoint.\n3. Review bounded evidence with `ghosty fleet logs \u003Cworker-id>` and\n   `ghosty fleet artifacts \u003Cworker-id>`. Summarize artifact refs, not full\n   payloads.\n4. Classify the state before acting:\n   - `transient failure`: transport error, timeout, stale heartbeat, host\n     unavailable, or retryable provider\u002Fnetwork failure.\n   - `task failure`: worker completed the task but the result is wrong,\n     missing required artifacts, or reports a domain error.\n   - `verifier failure`: scorer\u002Fverifier failed or disagrees with the worker\n     result.\n   - `needs-human`: missing authority, unsafe secret boundary, destructive\n     action, repeated restart exhaustion, ambiguous product decision, or\n     conflict between artifacts and verifier.\n5. Choose one typed action:\n   - transient and retry budget remains: `ghosty fleet restart \u003Cworker-id>`.\n   - transient but unsafe to retry: draft escalation and mark needs-human.\n   - task failure: preserve artifacts, summarize the failure, and avoid restart\n     unless the task spec says retrying can produce new evidence.\n   - verifier failure: inspect scorer inputs and artifacts, then escalate if the\n     verifier cannot be corrected through a typed action.\n   - needs-human: do not restart automatically; draft a concise escalation.\n6. Record the result in the response: classification, action taken or drafted,\n   evidence commands, artifact refs, and next owner.\n\n## Restart vs Escalate\n\nRestart only when all of these are true:\n\n- the failure is likely transient,\n- the task is idempotent or the run policy allows retry,\n- retry budget remains,\n- no secret, permission, or destructive action boundary is involved, and\n- the previous attempt produced enough receipt data to explain the restart.\n\nEscalate when any of these are true:\n\n- restart budget is exhausted,\n- the worker requests secrets or new authority,\n- artifacts indicate data loss, corruption, or destructive side effects,\n- the verifier and task result conflict in a way you cannot resolve from typed\n  evidence,\n- the same failure repeats after a restart, or\n- a human product or release decision is required.\n\n## Safe Escalation Draft\n\nUse this shape for Slack\u002FPagerDuty drafts. Keep logs to three short lines or an\nartifact ref.\n\n```text\nGhosty fleet needs attention\nRun: \u003Crun-id>\nWorker: \u003Cworker-id>\nTask: \u003Ctask-id or unknown>\nClassification: \u003Ctransient failure | task failure | verifier failure | needs-human>\nReason: \u003Cone sentence, no secrets>\nLatest typed evidence: ghosty fleet inspect \u003Cworker-id>; ghosty fleet artifacts \u003Cworker-id>\nSafe log excerpt: \u003C3 lines max or \"see artifact \u003Cref>\">\nRequested decision: \u003Crestart approval | verifier review | task owner review | permission decision>\n```\n\n## Post-Run Receipt\n\nEnd every fleet-manager response with a compact receipt:\n\n```text\nFleet receipt\nRun: \u003Crun-id>\nWorkers checked: \u003Ccount\u002Flist>\nClassification: \u003Cstate>\nAction: \u003Crestart\u002Finterrupt\u002Fstop\u002Fescalation draft\u002Fno-op>\nLedger expectation: \u003Ctyped action should be recorded | draft only, no send>\nArtifacts reviewed: \u003Crefs>\nFollow-up owner: \u003Cmanager | task owner | human>\n```\n","# Fleet Manager\n\nUse this skill when acting as a manager agent for Ghosty Agent Fleet runs.\nYour job is to classify worker state, choose the narrowest safe typed action,\nand leave a ledgered receipt or a safe escalation draft.\n\n## Authority Boundary\n\n- Prefer typed fleet surfaces over shell spelunking: `ghosty fleet status`,\n  `inspect`, `logs`, `artifacts`, `interrupt`, `restart`, `stop`, and the\n  Runtime API fleet endpoints.\n- Do not read `.ghosty\u002Ffleet.jsonl`, host logs, or remote files directly\n  unless the typed command or API is missing required evidence.\n- Do not send Slack, webhook, PagerDuty, email, or chat messages unless the\n  user or run config explicitly authorizes sending. Draft the message instead.\n- Never include secrets, tokens, webhook URLs, routing keys, full prompts, or\n  oversized logs in a summary or escalation.\n\n## Triage Loop\n\n1. Identify the run and worker from the user request, run receipt, or fleet\n   status output. If no worker is named, start with `ghosty fleet status`.\n2. Inspect the worker with `ghosty fleet inspect \u003Cworker-id>` or the matching\n   Runtime API worker endpoint.\n3. Review bounded evidence with `ghosty fleet logs \u003Cworker-id>` and\n   `ghosty fleet artifacts \u003Cworker-id>`. Summarize artifact refs, not full\n   payloads.\n4. Classify the state before acting:\n   - `transient failure`: transport error, timeout, stale heartbeat, host\n     unavailable, or retryable provider\u002Fnetwork failure.\n   - `task failure`: worker completed the task but the result is wrong,\n     missing required artifacts, or reports a domain error.\n   - `verifier failure`: scorer\u002Fverifier failed or disagrees with the worker\n     result.\n   - `needs-human`: missing authority, unsafe secret boundary, destructive\n     action, repeated restart exhaustion, ambiguous product decision, or\n     conflict between artifacts and verifier.\n5. Choose one typed action:\n   - transient and retry budget remains: `ghosty fleet restart \u003Cworker-id>`.\n   - transient but unsafe to retry: draft escalation and mark needs-human.\n   - task failure: preserve artifacts, summarize the failure, and avoid restart\n     unless the task spec says retrying can produce new evidence.\n   - verifier failure: inspect scorer inputs and artifacts, then escalate if the\n     verifier cannot be corrected through a typed action.\n   - needs-human: do not restart automatically; draft a concise escalation.\n6. Record the result in the response: classification, action taken or drafted,\n   evidence commands, artifact refs, and next owner.\n\n## Restart vs Escalate\n\nRestart only when all of these are true:\n\n- the failure is likely transient,\n- the task is idempotent or the run policy allows retry,\n- retry budget remains,\n- no secret, permission, or destructive action boundary is involved, and\n- the previous attempt produced enough receipt data to explain the restart.\n\nEscalate when any of these are true:\n\n- restart budget is exhausted,\n- the worker requests secrets or new authority,\n- artifacts indicate data loss, corruption, or destructive side effects,\n- the verifier and task result conflict in a way you cannot resolve from typed\n  evidence,\n- the same failure repeats after a restart, or\n- a human product or release decision is required.\n\n## Safe Escalation Draft\n\nUse this shape for Slack\u002FPagerDuty drafts. Keep logs to three short lines or an\nartifact ref.\n\n```text\nGhosty fleet needs attention\nRun: \u003Crun-id>\nWorker: \u003Cworker-id>\nTask: \u003Ctask-id or unknown>\nClassification: \u003Ctransient failure | task failure | verifier failure | needs-human>\nReason: \u003Cone sentence, no secrets>\nLatest typed evidence: ghosty fleet inspect \u003Cworker-id>; ghosty fleet artifacts \u003Cworker-id>\nSafe log excerpt: \u003C3 lines max or \"see artifact \u003Cref>\">\nRequested decision: \u003Crestart approval | verifier review | task owner review | permission decision>\n```\n\n## Post-Run Receipt\n\nEnd every fleet-manager response with a compact receipt:\n\n```text\nFleet receipt\nRun: \u003Crun-id>\nWorkers checked: \u003Ccount\u002Flist>\nClassification: \u003Cstate>\nAction: \u003Crestart\u002Finterrupt\u002Fstop\u002Fescalation draft\u002Fno-op>\nLedger expectation: \u003Ctyped action should be recorded | draft only, no send>\nArtifacts reviewed: \u003Crefs>\nFollow-up owner: \u003Cmanager | task owner | human>\n```\n",{"name":5,"metadata":23,"description":7},{"short-description":24},"Triage Ghosty Agent Fleet runs","main","c025ba29f3a55734ac05ff16ba677765a2845cfb","crates\u002Ftui\u002Fassets\u002Fskills\u002Ffleet-manager",0,"https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode","https:\u002F\u002Fgithub.com\u002Fblissito","https:\u002F\u002Ffixtergeek.com","Off-grid developer writing code since 1999 🤓","2026-11-01 19:47:50",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add blissito\u002Fghostycode --skill fleet-manager",{"tool":40,"command":41},"bunx","bunx skills add blissito\u002Fghostycode --skill fleet-manager",{"tool":43,"command":44},"pnpm","pnpm dlx skills add blissito\u002Fghostycode --skill fleet-manager","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fcrates\u002Ftui\u002Fassets\u002Fskills\u002Ffleet-manager\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":18},"01M75CGWC29615C661HTEGXVTH","gh-treasure-hunt","Hunt the issue\u002FPR queue for highest value-over-risk wins: clean focused community PRs, already-implemented issues to close, safe quick-fixes.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-treasure-hunt\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":18},"01M75CGTGM88WWQKNKB7G3F688","gh-find-prs","Survey open GhostyCode PRs and triage each for mergeability and disposition against the real landing branch.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-find-prs\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":18},"01M75CGQ1REETJFS53P5V9KH3E","gh-file-issue","Use when filing a new GhostyCode GitHub issue: turn a bug or idea into a well-formed, actionable issue with repro, acceptance criteria, labels, and milestone.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-file-issue\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":18},"01M75CGK915XTS0EG9F4NSVZ6G","gh-credit-harvest","Harvest one community PR into a release branch with authorship and credit preserved, verified green, and a warm thank-you.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-credit-harvest\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":18},"01M75CGH3767G7218T7FV4TGM8","gh-compile-issues","Triage N GitHub issues into a coverage matrix: fetch each, check current code, classify already-done\u002Fquick-fix\u002Fdesign\u002Fdefer with cited evidence.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-compile-issues\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":18},"01M75CGF7A38RNP14045E43Q1R","gh-close-issues","Close resolved GhostyCode issues only after verifying the landed commit\u002Fbehavior, with a positive crediting comment; never from title alone.","https:\u002F\u002Fgithub.com\u002Fblissito\u002Fghostycode\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fgh-close-issues\u002FSKILL.md",[],[]]