[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-aws-labs-cao-session-liveness":3,"mdc--x3s5fx-key":36,"related-org-aws-labs-cao-session-liveness":731,"related-repo-aws-labs-cao-session-liveness":909},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":12,"stars":26,"repoUrl":27,"updatedAt":28,"license":29,"forks":30,"topics":31,"repo":32,"sourceUrl":34,"mdContent":35},"cao-session-liveness","verify CAO session liveness","Verify whether a CAO session is actually alive and what it really said, before reporting progress or completion to a user. Use alongside cao-session-management whenever you launch, poll, or report on a CAO session — especially when a session appears stalled, a send times out, or a status value looks inconsistent with the output.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},"aws-labs","AWS Labs","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Faws-labs.png","awslabs",[13,17,20,23],{"name":14,"slug":15,"type":16},"Operations","operations","tag",{"name":18,"slug":19,"type":16},"Monitoring","monitoring",{"name":21,"slug":22,"type":16},"AWS","aws",{"name":24,"slug":25,"type":16},"Debugging","debugging",871,"https:\u002F\u002Fgithub.com\u002Fawslabs\u002Fcli-agent-orchestrator","2026-08-23T04:00:13.076812",null,164,[],{"repoUrl":27,"stars":26,"forks":30,"topics":33,"description":29},[],"https:\u002F\u002Fgithub.com\u002Fawslabs\u002Fcli-agent-orchestrator\u002Ftree\u002FHEAD\u002Fskills\u002Fcao-session-liveness","---\nname: cao-session-liveness\ndescription: Verify whether a CAO session is actually alive and what it really\n  said, before reporting progress or completion to a user. Use alongside\n  cao-session-management whenever you launch, poll, or report on a CAO session —\n  especially when a session appears stalled, a send times out, or a status value\n  looks inconsistent with the output.\n---\n\n# CAO Session Liveness\n\nCompanion to `cao-session-management`, which covers the mechanics of launching\nand messaging sessions. This skill covers a single question that mechanics alone\ncannot answer: **is the session actually alive, and is the status telling me the\ntruth?**\n\n## Why this matters\n\nEvery CAO provider infers agent state by pattern-matching the rendered terminal\nscreen. There is no structured protocol between CAO and the provider CLI. A\nprovider that has exited, crashed, or stalled on an unanswerable dialog can\nleave a screen that still matches an `idle` or `processing` pattern.\n\nThe consequence is specific and it is the failure this skill exists to prevent:\n**reporting progress on a session that is already dead.**\n\n## The two-signal rule\n\nNever report readiness, progress, or completion from a status value alone.\nAlways corroborate with output before you tell a user anything:\n\n1. Read the status (`get_terminal_status`, or `cao session status SESSION`).\n2. Read the output tail (`read_session_output` \u002F `get_terminal_output`, or\n   `cao session status SESSION --json` and inspect `last_output`).\n3. If the two disagree, **the output wins.**\n\nA status of `idle` with an output tail showing a shell prompt means the CLI\nexited. Report the session as dead, not as ready.\n\n## Dead-session discriminators\n\nTreat any of the following in the output tail as proof the provider is no longer\nrunning, regardless of the reported status:\n\n| Signal | Means |\n|---|---|\n| `Session ended.` \u002F `Resume with: \u003Ccli> --resume-id ...` | The CLI exited on its own |\n| `error: Conflicting options:` or a usage\u002Fhelp banner | The CLI rejected its launch flags and never started |\n| `API Error (...)`, `400`, or a model\u002Fauth failure | The provider started but cannot reach a model |\n| A bare shell prompt with a directory and timestamp, no agent chrome | The pane fell back to the shell |\n| An output read that fails with an extraction error | No response boundary on screen; corroborate before trusting |\n\nA session parked in `waiting_user_answer` that never advances is usually stalled\non a dialog nothing will answer. Treat it as dead weight, report it to the user,\nand do not silently kill it.\n\n### Not a dead session: a finished handoff worker\n\nA blocking `handoff` tears its worker down once it returns. The worker terminal\nID the conductor reports was valid **during** the call and is gone afterwards, so\nquerying it later is expected to fail:\n\n- `get_terminal_status` \u002F `GET \u002Fterminals\u002F\u003Cid>` returns not-found\n- `cao session status SESSION --workers` lists no workers\n\nNeither is evidence the conductor invented the delegation. Confirm a handoff from\nthe **conductor's own transcript** — a full-mode output read showing the\n`handoff` tool call, its `agent_profile`, and the returned output — not from the\nterminal registry. Only a non-blocking `assign` leaves a worker alive to query.\n\nDo not accuse a conductor of fabricating a delegation on the strength of a\nmissing terminal alone.\n\n## Verify a provider before depending on it\n\nProvider reliability varies, is version-sensitive, and changes as upstream CLIs\nrelease new dialogs and flags. Do not assume; verify once per environment:\n\n1. Launch a throwaway session in a scratch directory.\n2. Apply the two-signal rule.\n3. Send a trivial task with a short timeout and confirm output returns.\n4. Shut the session down.\n\nKnown reliability characteristics, as context for interpreting what you see:\n\n| Provider | Detection basis | What to watch for |\n|---|---|---|\n| `kiro_cli` | Version-specific prompt, credits, and separator patterns | New startup dialogs that default to a decline option; flag combinations the installed CLI rejects |\n| `hermes` | Idle timer stable across repeated polls | Custom themes break prompt matching; slowest to confirm completion. Patterns are overridable by environment variable |\n| `opencode_cli` | Alt-screen TUI completion marker | Scrollback is roughly one viewport; a long single response can lose its own top and fail extraction |\n| `claude_code`, `codex` | Rendered-screen detection | Generally stable headless; still apply the two-signal rule |\n\nIf a provider fails to launch headlessly, report the exact signature to the user\nand offer a different provider. Do not retry the same launch repeatedly — a flag\nrejection or a declining dialog will fail identically every time.\n\n## Interpreting a send that does not return\n\n- A **timeout is not a failure.** The agent is still working; the caller stopped\n  waiting. Say so, and check again later.\n- **Never re-send a task after a timeout.** The original may still be running,\n  and a duplicate risks conflicting work in the same directory.\n- A **busy terminal refuses input.** Wait for `idle` or `completed`; do not force.\n- An **async send returns nothing by design.** Poll afterwards, applying the\n  two-signal rule.\n\n## Record what each session is for\n\nCAO stores a session's name, not its purpose. An inventory of live sessions\ncannot tell you which is safe to touch.\n\nKeep a short registry outside CAO — one line per session you launch: name,\nprovider, working directory, purpose, date. Update it on launch and on\nshutdown, and read it before answering any question about what a session is\ndoing or before acting on one.\n\n## Do not act on sessions you did not launch\n\nLong-running sessions may hold real, unrecoverable work. Reads are always safe.\nBefore sending to or shutting down a session you did not start yourself, ask the\nuser first. Never issue a shutdown that targets all sessions at once.\n\n## Related\n\n- [cao-session-management](..\u002Fcao-session-management\u002FSKILL.md) — launching,\n  messaging, and worker communication mechanics\n",{"data":37,"body":38},{"name":4,"description":6},{"type":39,"children":40},"root",[41,49,70,77,98,108,114,119,191,203,209,214,338,351,358,378,410,445,450,456,461,484,489,611,616,622,684,690,695,700,706,711,717],{"type":42,"tag":43,"props":44,"children":45},"element","h1",{"id":4},[46],{"type":47,"value":48},"text","CAO Session Liveness",{"type":42,"tag":50,"props":51,"children":52},"p",{},[53,55,62,64],{"type":47,"value":54},"Companion to ",{"type":42,"tag":56,"props":57,"children":59},"code",{"className":58},[],[60],{"type":47,"value":61},"cao-session-management",{"type":47,"value":63},", which covers the mechanics of launching\nand messaging sessions. This skill covers a single question that mechanics alone\ncannot answer: ",{"type":42,"tag":65,"props":66,"children":67},"strong",{},[68],{"type":47,"value":69},"is the session actually alive, and is the status telling me the\ntruth?",{"type":42,"tag":71,"props":72,"children":74},"h2",{"id":73},"why-this-matters",[75],{"type":47,"value":76},"Why this matters",{"type":42,"tag":50,"props":78,"children":79},{},[80,82,88,90,96],{"type":47,"value":81},"Every CAO provider infers agent state by pattern-matching the rendered terminal\nscreen. There is no structured protocol between CAO and the provider CLI. A\nprovider that has exited, crashed, or stalled on an unanswerable dialog can\nleave a screen that still matches an ",{"type":42,"tag":56,"props":83,"children":85},{"className":84},[],[86],{"type":47,"value":87},"idle",{"type":47,"value":89}," or ",{"type":42,"tag":56,"props":91,"children":93},{"className":92},[],[94],{"type":47,"value":95},"processing",{"type":47,"value":97}," pattern.",{"type":42,"tag":50,"props":99,"children":100},{},[101,103],{"type":47,"value":102},"The consequence is specific and it is the failure this skill exists to prevent:\n",{"type":42,"tag":65,"props":104,"children":105},{},[106],{"type":47,"value":107},"reporting progress on a session that is already dead.",{"type":42,"tag":71,"props":109,"children":111},{"id":110},"the-two-signal-rule",[112],{"type":47,"value":113},"The two-signal rule",{"type":42,"tag":50,"props":115,"children":116},{},[117],{"type":47,"value":118},"Never report readiness, progress, or completion from a status value alone.\nAlways corroborate with output before you tell a user anything:",{"type":42,"tag":120,"props":121,"children":122},"ol",{},[123,145,181],{"type":42,"tag":124,"props":125,"children":126},"li",{},[127,129,135,137,143],{"type":47,"value":128},"Read the status (",{"type":42,"tag":56,"props":130,"children":132},{"className":131},[],[133],{"type":47,"value":134},"get_terminal_status",{"type":47,"value":136},", or ",{"type":42,"tag":56,"props":138,"children":140},{"className":139},[],[141],{"type":47,"value":142},"cao session status SESSION",{"type":47,"value":144},").",{"type":42,"tag":124,"props":146,"children":147},{},[148,150,156,158,164,166,172,174,180],{"type":47,"value":149},"Read the output tail (",{"type":42,"tag":56,"props":151,"children":153},{"className":152},[],[154],{"type":47,"value":155},"read_session_output",{"type":47,"value":157}," \u002F ",{"type":42,"tag":56,"props":159,"children":161},{"className":160},[],[162],{"type":47,"value":163},"get_terminal_output",{"type":47,"value":165},", or\n",{"type":42,"tag":56,"props":167,"children":169},{"className":168},[],[170],{"type":47,"value":171},"cao session status SESSION --json",{"type":47,"value":173}," and inspect ",{"type":42,"tag":56,"props":175,"children":177},{"className":176},[],[178],{"type":47,"value":179},"last_output",{"type":47,"value":144},{"type":42,"tag":124,"props":182,"children":183},{},[184,186],{"type":47,"value":185},"If the two disagree, ",{"type":42,"tag":65,"props":187,"children":188},{},[189],{"type":47,"value":190},"the output wins.",{"type":42,"tag":50,"props":192,"children":193},{},[194,196,201],{"type":47,"value":195},"A status of ",{"type":42,"tag":56,"props":197,"children":199},{"className":198},[],[200],{"type":47,"value":87},{"type":47,"value":202}," with an output tail showing a shell prompt means the CLI\nexited. Report the session as dead, not as ready.",{"type":42,"tag":71,"props":204,"children":206},{"id":205},"dead-session-discriminators",[207],{"type":47,"value":208},"Dead-session discriminators",{"type":42,"tag":50,"props":210,"children":211},{},[212],{"type":47,"value":213},"Treat any of the following in the output tail as proof the provider is no longer\nrunning, regardless of the reported status:",{"type":42,"tag":215,"props":216,"children":217},"table",{},[218,237],{"type":42,"tag":219,"props":220,"children":221},"thead",{},[222],{"type":42,"tag":223,"props":224,"children":225},"tr",{},[226,232],{"type":42,"tag":227,"props":228,"children":229},"th",{},[230],{"type":47,"value":231},"Signal",{"type":42,"tag":227,"props":233,"children":234},{},[235],{"type":47,"value":236},"Means",{"type":42,"tag":238,"props":239,"children":240},"tbody",{},[241,266,285,312,325],{"type":42,"tag":223,"props":242,"children":243},{},[244,261],{"type":42,"tag":245,"props":246,"children":247},"td",{},[248,254,255],{"type":42,"tag":56,"props":249,"children":251},{"className":250},[],[252],{"type":47,"value":253},"Session ended.",{"type":47,"value":157},{"type":42,"tag":56,"props":256,"children":258},{"className":257},[],[259],{"type":47,"value":260},"Resume with: \u003Ccli> --resume-id ...",{"type":42,"tag":245,"props":262,"children":263},{},[264],{"type":47,"value":265},"The CLI exited on its own",{"type":42,"tag":223,"props":267,"children":268},{},[269,280],{"type":42,"tag":245,"props":270,"children":271},{},[272,278],{"type":42,"tag":56,"props":273,"children":275},{"className":274},[],[276],{"type":47,"value":277},"error: Conflicting options:",{"type":47,"value":279}," or a usage\u002Fhelp banner",{"type":42,"tag":245,"props":281,"children":282},{},[283],{"type":47,"value":284},"The CLI rejected its launch flags and never started",{"type":42,"tag":223,"props":286,"children":287},{},[288,307],{"type":42,"tag":245,"props":289,"children":290},{},[291,297,299,305],{"type":42,"tag":56,"props":292,"children":294},{"className":293},[],[295],{"type":47,"value":296},"API Error (...)",{"type":47,"value":298},", ",{"type":42,"tag":56,"props":300,"children":302},{"className":301},[],[303],{"type":47,"value":304},"400",{"type":47,"value":306},", or a model\u002Fauth failure",{"type":42,"tag":245,"props":308,"children":309},{},[310],{"type":47,"value":311},"The provider started but cannot reach a model",{"type":42,"tag":223,"props":313,"children":314},{},[315,320],{"type":42,"tag":245,"props":316,"children":317},{},[318],{"type":47,"value":319},"A bare shell prompt with a directory and timestamp, no agent chrome",{"type":42,"tag":245,"props":321,"children":322},{},[323],{"type":47,"value":324},"The pane fell back to the shell",{"type":42,"tag":223,"props":326,"children":327},{},[328,333],{"type":42,"tag":245,"props":329,"children":330},{},[331],{"type":47,"value":332},"An output read that fails with an extraction error",{"type":42,"tag":245,"props":334,"children":335},{},[336],{"type":47,"value":337},"No response boundary on screen; corroborate before trusting",{"type":42,"tag":50,"props":339,"children":340},{},[341,343,349],{"type":47,"value":342},"A session parked in ",{"type":42,"tag":56,"props":344,"children":346},{"className":345},[],[347],{"type":47,"value":348},"waiting_user_answer",{"type":47,"value":350}," that never advances is usually stalled\non a dialog nothing will answer. Treat it as dead weight, report it to the user,\nand do not silently kill it.",{"type":42,"tag":352,"props":353,"children":355},"h3",{"id":354},"not-a-dead-session-a-finished-handoff-worker",[356],{"type":47,"value":357},"Not a dead session: a finished handoff worker",{"type":42,"tag":50,"props":359,"children":360},{},[361,363,369,371,376],{"type":47,"value":362},"A blocking ",{"type":42,"tag":56,"props":364,"children":366},{"className":365},[],[367],{"type":47,"value":368},"handoff",{"type":47,"value":370}," tears its worker down once it returns. The worker terminal\nID the conductor reports was valid ",{"type":42,"tag":65,"props":372,"children":373},{},[374],{"type":47,"value":375},"during",{"type":47,"value":377}," the call and is gone afterwards, so\nquerying it later is expected to fail:",{"type":42,"tag":379,"props":380,"children":381},"ul",{},[382,399],{"type":42,"tag":124,"props":383,"children":384},{},[385,390,391,397],{"type":42,"tag":56,"props":386,"children":388},{"className":387},[],[389],{"type":47,"value":134},{"type":47,"value":157},{"type":42,"tag":56,"props":392,"children":394},{"className":393},[],[395],{"type":47,"value":396},"GET \u002Fterminals\u002F\u003Cid>",{"type":47,"value":398}," returns not-found",{"type":42,"tag":124,"props":400,"children":401},{},[402,408],{"type":42,"tag":56,"props":403,"children":405},{"className":404},[],[406],{"type":47,"value":407},"cao session status SESSION --workers",{"type":47,"value":409}," lists no workers",{"type":42,"tag":50,"props":411,"children":412},{},[413,415,420,422,427,429,435,437,443],{"type":47,"value":414},"Neither is evidence the conductor invented the delegation. Confirm a handoff from\nthe ",{"type":42,"tag":65,"props":416,"children":417},{},[418],{"type":47,"value":419},"conductor's own transcript",{"type":47,"value":421}," — a full-mode output read showing the\n",{"type":42,"tag":56,"props":423,"children":425},{"className":424},[],[426],{"type":47,"value":368},{"type":47,"value":428}," tool call, its ",{"type":42,"tag":56,"props":430,"children":432},{"className":431},[],[433],{"type":47,"value":434},"agent_profile",{"type":47,"value":436},", and the returned output — not from the\nterminal registry. Only a non-blocking ",{"type":42,"tag":56,"props":438,"children":440},{"className":439},[],[441],{"type":47,"value":442},"assign",{"type":47,"value":444}," leaves a worker alive to query.",{"type":42,"tag":50,"props":446,"children":447},{},[448],{"type":47,"value":449},"Do not accuse a conductor of fabricating a delegation on the strength of a\nmissing terminal alone.",{"type":42,"tag":71,"props":451,"children":453},{"id":452},"verify-a-provider-before-depending-on-it",[454],{"type":47,"value":455},"Verify a provider before depending on it",{"type":42,"tag":50,"props":457,"children":458},{},[459],{"type":47,"value":460},"Provider reliability varies, is version-sensitive, and changes as upstream CLIs\nrelease new dialogs and flags. Do not assume; verify once per environment:",{"type":42,"tag":120,"props":462,"children":463},{},[464,469,474,479],{"type":42,"tag":124,"props":465,"children":466},{},[467],{"type":47,"value":468},"Launch a throwaway session in a scratch directory.",{"type":42,"tag":124,"props":470,"children":471},{},[472],{"type":47,"value":473},"Apply the two-signal rule.",{"type":42,"tag":124,"props":475,"children":476},{},[477],{"type":47,"value":478},"Send a trivial task with a short timeout and confirm output returns.",{"type":42,"tag":124,"props":480,"children":481},{},[482],{"type":47,"value":483},"Shut the session down.",{"type":42,"tag":50,"props":485,"children":486},{},[487],{"type":47,"value":488},"Known reliability characteristics, as context for interpreting what you see:",{"type":42,"tag":215,"props":490,"children":491},{},[492,513],{"type":42,"tag":219,"props":493,"children":494},{},[495],{"type":42,"tag":223,"props":496,"children":497},{},[498,503,508],{"type":42,"tag":227,"props":499,"children":500},{},[501],{"type":47,"value":502},"Provider",{"type":42,"tag":227,"props":504,"children":505},{},[506],{"type":47,"value":507},"Detection basis",{"type":42,"tag":227,"props":509,"children":510},{},[511],{"type":47,"value":512},"What to watch for",{"type":42,"tag":238,"props":514,"children":515},{},[516,538,560,582],{"type":42,"tag":223,"props":517,"children":518},{},[519,528,533],{"type":42,"tag":245,"props":520,"children":521},{},[522],{"type":42,"tag":56,"props":523,"children":525},{"className":524},[],[526],{"type":47,"value":527},"kiro_cli",{"type":42,"tag":245,"props":529,"children":530},{},[531],{"type":47,"value":532},"Version-specific prompt, credits, and separator patterns",{"type":42,"tag":245,"props":534,"children":535},{},[536],{"type":47,"value":537},"New startup dialogs that default to a decline option; flag combinations the installed CLI rejects",{"type":42,"tag":223,"props":539,"children":540},{},[541,550,555],{"type":42,"tag":245,"props":542,"children":543},{},[544],{"type":42,"tag":56,"props":545,"children":547},{"className":546},[],[548],{"type":47,"value":549},"hermes",{"type":42,"tag":245,"props":551,"children":552},{},[553],{"type":47,"value":554},"Idle timer stable across repeated polls",{"type":42,"tag":245,"props":556,"children":557},{},[558],{"type":47,"value":559},"Custom themes break prompt matching; slowest to confirm completion. Patterns are overridable by environment variable",{"type":42,"tag":223,"props":561,"children":562},{},[563,572,577],{"type":42,"tag":245,"props":564,"children":565},{},[566],{"type":42,"tag":56,"props":567,"children":569},{"className":568},[],[570],{"type":47,"value":571},"opencode_cli",{"type":42,"tag":245,"props":573,"children":574},{},[575],{"type":47,"value":576},"Alt-screen TUI completion marker",{"type":42,"tag":245,"props":578,"children":579},{},[580],{"type":47,"value":581},"Scrollback is roughly one viewport; a long single response can lose its own top and fail extraction",{"type":42,"tag":223,"props":583,"children":584},{},[585,601,606],{"type":42,"tag":245,"props":586,"children":587},{},[588,594,595],{"type":42,"tag":56,"props":589,"children":591},{"className":590},[],[592],{"type":47,"value":593},"claude_code",{"type":47,"value":298},{"type":42,"tag":56,"props":596,"children":598},{"className":597},[],[599],{"type":47,"value":600},"codex",{"type":42,"tag":245,"props":602,"children":603},{},[604],{"type":47,"value":605},"Rendered-screen detection",{"type":42,"tag":245,"props":607,"children":608},{},[609],{"type":47,"value":610},"Generally stable headless; still apply the two-signal rule",{"type":42,"tag":50,"props":612,"children":613},{},[614],{"type":47,"value":615},"If a provider fails to launch headlessly, report the exact signature to the user\nand offer a different provider. Do not retry the same launch repeatedly — a flag\nrejection or a declining dialog will fail identically every time.",{"type":42,"tag":71,"props":617,"children":619},{"id":618},"interpreting-a-send-that-does-not-return",[620],{"type":47,"value":621},"Interpreting a send that does not return",{"type":42,"tag":379,"props":623,"children":624},{},[625,637,647,672],{"type":42,"tag":124,"props":626,"children":627},{},[628,630,635],{"type":47,"value":629},"A ",{"type":42,"tag":65,"props":631,"children":632},{},[633],{"type":47,"value":634},"timeout is not a failure.",{"type":47,"value":636}," The agent is still working; the caller stopped\nwaiting. Say so, and check again later.",{"type":42,"tag":124,"props":638,"children":639},{},[640,645],{"type":42,"tag":65,"props":641,"children":642},{},[643],{"type":47,"value":644},"Never re-send a task after a timeout.",{"type":47,"value":646}," The original may still be running,\nand a duplicate risks conflicting work in the same directory.",{"type":42,"tag":124,"props":648,"children":649},{},[650,651,656,658,663,664,670],{"type":47,"value":629},{"type":42,"tag":65,"props":652,"children":653},{},[654],{"type":47,"value":655},"busy terminal refuses input.",{"type":47,"value":657}," Wait for ",{"type":42,"tag":56,"props":659,"children":661},{"className":660},[],[662],{"type":47,"value":87},{"type":47,"value":89},{"type":42,"tag":56,"props":665,"children":667},{"className":666},[],[668],{"type":47,"value":669},"completed",{"type":47,"value":671},"; do not force.",{"type":42,"tag":124,"props":673,"children":674},{},[675,677,682],{"type":47,"value":676},"An ",{"type":42,"tag":65,"props":678,"children":679},{},[680],{"type":47,"value":681},"async send returns nothing by design.",{"type":47,"value":683}," Poll afterwards, applying the\ntwo-signal rule.",{"type":42,"tag":71,"props":685,"children":687},{"id":686},"record-what-each-session-is-for",[688],{"type":47,"value":689},"Record what each session is for",{"type":42,"tag":50,"props":691,"children":692},{},[693],{"type":47,"value":694},"CAO stores a session's name, not its purpose. An inventory of live sessions\ncannot tell you which is safe to touch.",{"type":42,"tag":50,"props":696,"children":697},{},[698],{"type":47,"value":699},"Keep a short registry outside CAO — one line per session you launch: name,\nprovider, working directory, purpose, date. Update it on launch and on\nshutdown, and read it before answering any question about what a session is\ndoing or before acting on one.",{"type":42,"tag":71,"props":701,"children":703},{"id":702},"do-not-act-on-sessions-you-did-not-launch",[704],{"type":47,"value":705},"Do not act on sessions you did not launch",{"type":42,"tag":50,"props":707,"children":708},{},[709],{"type":47,"value":710},"Long-running sessions may hold real, unrecoverable work. Reads are always safe.\nBefore sending to or shutting down a session you did not start yourself, ask the\nuser first. Never issue a shutdown that targets all sessions at once.",{"type":42,"tag":71,"props":712,"children":714},{"id":713},"related",[715],{"type":47,"value":716},"Related",{"type":42,"tag":379,"props":718,"children":719},{},[720],{"type":42,"tag":124,"props":721,"children":722},{},[723,729],{"type":42,"tag":724,"props":725,"children":727},"a",{"href":726},"..\u002Fcao-session-management\u002FSKILL.md",[728],{"type":47,"value":61},{"type":47,"value":730}," — launching,\nmessaging, and worker communication mechanics",{"items":732,"total":908},[733,750,771,781,794,807,817,827,848,863,878,893],{"slug":734,"name":734,"fn":735,"description":736,"org":737,"tags":738,"stars":747,"repoUrl":748,"updatedAt":749},"agentcore-investigation","investigate Bedrock AgentCore runtime sessions","Investigate Bedrock AgentCore runtime sessions via CloudWatch Logs Insights — resolve session\u002Ftrace IDs, query OTEL spans, filter noise, build timelines. Use when debugging AgentCore agent sessions, tracing tool calls, or analyzing latency.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[739,740,741,744],{"name":21,"slug":22,"type":16},{"name":24,"slug":25,"type":16},{"name":742,"slug":743,"type":16},"Logs","logs",{"name":745,"slug":746,"type":16},"Observability","observability",9427,"https:\u002F\u002Fgithub.com\u002Fawslabs\u002Fmcp","2026-07-12T08:37:22.601527",{"slug":751,"name":752,"fn":753,"description":754,"org":755,"tags":756,"stars":747,"repoUrl":748,"updatedAt":770},"amazon-aurora-dsql","amazon aurora dsql","build applications with Aurora DSQL","Build with Aurora DSQL — manage schemas, execute queries, handle migrations, diagnose query plans, load data, and develop applications with a serverless, distributed SQL database. Covers IAM auth, multi-tenant patterns, MySQL-to-DSQL and PostgreSQL-to-DSQL schema conversion, FK replacement code generation, OCC retry patterns, ORM migration (Django\u002FHibernate\u002FRails), DDL operations, query plan explainability, SQL compatibility validation, and bulk data loading. Triggers on phrases like: DSQL, Aurora DSQL, create DSQL table, DSQL schema, migrate to DSQL, distributed SQL database, serverless PostgreSQL-compatible database, DSQL query plan, DSQL EXPLAIN ANALYZE, why is my DSQL query slow, DSQL foreign key, DSQL OCC retry, DSQL multi-region, load into DSQL, load CSV into DSQL, bulk load DSQL, aurora-dsql-loader.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[757,760,761,764,767],{"name":758,"slug":759,"type":16},"Aurora","aurora",{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},"Database","database",{"name":765,"slug":766,"type":16},"Serverless","serverless",{"name":768,"slug":769,"type":16},"SQL","sql","2026-08-04T05:35:10.770847",{"slug":772,"name":773,"fn":753,"description":754,"org":774,"tags":775,"stars":747,"repoUrl":748,"updatedAt":780},"aurora-dsql","aurora dsql",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[776,777,778,779],{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},{"name":765,"slug":766,"type":16},{"name":768,"slug":769,"type":16},"2026-08-04T05:35:05.694395",{"slug":782,"name":783,"fn":753,"description":754,"org":784,"tags":785,"stars":747,"repoUrl":748,"updatedAt":793},"aws-dsql","aws dsql",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[786,787,788,791,792],{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},{"name":789,"slug":790,"type":16},"Migration","migration",{"name":765,"slug":766,"type":16},{"name":768,"slug":769,"type":16},"2026-08-04T05:35:08.749669",{"slug":795,"name":796,"fn":753,"description":754,"org":797,"tags":798,"stars":747,"repoUrl":748,"updatedAt":806},"distributed-postgres","distributed postgres",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[799,800,801,804,805],{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},{"name":802,"slug":803,"type":16},"PostgreSQL","postgresql",{"name":765,"slug":766,"type":16},{"name":768,"slug":769,"type":16},"2026-08-04T05:35:06.713102",{"slug":808,"name":809,"fn":753,"description":754,"org":810,"tags":811,"stars":747,"repoUrl":748,"updatedAt":816},"distributed-sql","distributed sql",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[812,813,814,815],{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},{"name":765,"slug":766,"type":16},{"name":768,"slug":769,"type":16},"2026-08-04T05:35:10.086942",{"slug":818,"name":818,"fn":753,"description":754,"org":819,"tags":820,"stars":747,"repoUrl":748,"updatedAt":826},"dsql",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[821,822,823,824,825],{"name":21,"slug":22,"type":16},{"name":762,"slug":763,"type":16},{"name":789,"slug":790,"type":16},{"name":765,"slug":766,"type":16},{"name":768,"slug":769,"type":16},"2026-08-04T05:35:07.751779",{"slug":828,"name":828,"fn":829,"description":830,"org":831,"tags":832,"stars":845,"repoUrl":846,"updatedAt":847},"cost-efficiency-analyzer","analyze cost efficiency and expenses","Analyzes cost structure, cost efficiency, and expense management from P&L data. Use when the user asks about costs, expenses, COGS, operating expenses, cost ratios, cost control, spending efficiency, margin compression from cost side, or wants to understand where money is going. Also use for \"are we spending too much\", \"cost breakdown\", \"expense analysis\", or \"how efficient are our operations\". NOT for revenue or top-line analysis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[833,836,839,842],{"name":834,"slug":835,"type":16},"Accounting","accounting",{"name":837,"slug":838,"type":16},"Analytics","analytics",{"name":840,"slug":841,"type":16},"Cost Optimization","cost-optimization",{"name":843,"slug":844,"type":16},"Finance","finance",3176,"https:\u002F\u002Fgithub.com\u002Fawslabs\u002Fagentcore-samples","2026-07-12T08:40:03.29555",{"slug":849,"name":849,"fn":850,"description":851,"org":852,"tags":853,"stars":845,"repoUrl":846,"updatedAt":862},"executive-financial-briefing","generate executive financial briefings","Generates a concise executive-level financial briefing or summary suitable for a CEO, CFO, or board presentation. Use when the user asks for a summary, briefing, executive summary, board update, financial overview, financial health check, or \"how is the business doing\". Covers the full P&L picture in one page. Also use for \"give me the highlights\", \"what do I need to know\", or \"quick financial update\".",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[854,855,856,859],{"name":21,"slug":22,"type":16},{"name":843,"slug":844,"type":16},{"name":857,"slug":858,"type":16},"Management","management",{"name":860,"slug":861,"type":16},"Reporting","reporting","2026-07-12T08:40:02.066471",{"slug":864,"name":864,"fn":865,"description":866,"org":867,"tags":868,"stars":845,"repoUrl":846,"updatedAt":877},"multi-quarter-trend-analysis","analyze multi-quarter financial trends","Analyzes financial trends across multiple quarters by comparing P&L metrics over time. Use when the user wants to see trends, patterns, trajectories, or directional movement across 3 or more quarters. Also use for \"how are we trending\", \"show me the trend\", \"track performance over time\", \"quarter over quarter comparison across all quarters\", or any multi-period longitudinal analysis.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[869,870,871,874],{"name":837,"slug":838,"type":16},{"name":843,"slug":844,"type":16},{"name":872,"slug":873,"type":16},"Financial Statements","financial-statements",{"name":875,"slug":876,"type":16},"Variance Analysis","variance-analysis","2026-07-12T08:40:00.79141",{"slug":879,"name":879,"fn":880,"description":881,"org":882,"tags":883,"stars":845,"repoUrl":846,"updatedAt":892},"pdf","process and manipulate PDF documents","Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text\u002Ftables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting\u002Fdecrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[884,887,890],{"name":885,"slug":886,"type":16},"Automation","automation",{"name":888,"slug":889,"type":16},"Documents","documents",{"name":891,"slug":879,"type":16},"PDF","2026-07-12T08:41:44.135656",{"slug":894,"name":894,"fn":895,"description":896,"org":897,"tags":898,"stars":845,"repoUrl":846,"updatedAt":907},"quarterly-kpi-calculator","calculate quarterly financial KPIs","Calculates quarterly financial KPIs from P&L data. P&L figures can be provided directly by the user or fetched from the financial data MCP server. Use when the user wants KPI calculations such as Gross Margin %, EBITDA Margin %, Operating Expense Ratio, or Revenue Growth % QoQ. Also use for quarterly performance review, P&L analysis, or interpreting financial ratios against benchmarks.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[899,900,903,904],{"name":834,"slug":835,"type":16},{"name":901,"slug":902,"type":16},"Data Analysis","data-analysis",{"name":843,"slug":844,"type":16},{"name":905,"slug":906,"type":16},"KPI","kpi","2026-07-12T08:39:59.54971",128,{"items":910,"total":1009},[911,930,945,957,971,984,998],{"slug":912,"name":912,"fn":913,"description":914,"org":915,"tags":916,"stars":26,"repoUrl":27,"updatedAt":929},"add-app-to-server","add interactive UI to MCP servers","This skill should be used when the user asks to \"add an app to my MCP server\", \"add UI to my MCP server\", \"add a view to my MCP tool\", \"enrich MCP tools with UI\", \"add interactive UI to existing server\", \"add MCP Apps to my server\", or needs to add interactive UI capabilities to an existing MCP server that already has tools. Provides guidance for analyzing existing tools and adding MCP Apps UI resources.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[917,920,923,926],{"name":918,"slug":919,"type":16},"Frontend","frontend",{"name":921,"slug":922,"type":16},"MCP","mcp",{"name":924,"slug":925,"type":16},"Plugin Development","plugin-development",{"name":927,"slug":928,"type":16},"UI Components","ui-components","2026-07-12T08:41:40.4008",{"slug":931,"name":931,"fn":932,"description":933,"org":934,"tags":935,"stars":26,"repoUrl":27,"updatedAt":944},"agui-author","emit interactive dashboard UI components","Author live dashboard UI from an agent via the `emit_ui` MCP tool. Emit one of six allow-listed components (approval_card, choice_prompt, diff_summary, progress, metric, agent_card) with JSON props and it renders in any AG-UI client watching the fleet. Use when you want the operator to see a decision, a diff, or a status readout instead of scrolling terminal text. Arbitrary HTML\u002Fmarkup is refused.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[936,939,942,943],{"name":937,"slug":938,"type":16},"Agents","agents",{"name":940,"slug":941,"type":16},"Dashboards","dashboards",{"name":921,"slug":922,"type":16},{"name":927,"slug":928,"type":16},"2026-07-22T05:35:50.851108",{"slug":946,"name":946,"fn":947,"description":948,"org":949,"tags":950,"stars":26,"repoUrl":27,"updatedAt":956},"cao-agent-routing","route tasks to appropriate CAO agents","Find and select the best installed CAO agent profile for a task before delegating with assign or handoff. Use when a supervisor needs to route coding, documentation, infrastructure, review, research, or other specialist work and the user has not already chosen an agent profile.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[951,952,953],{"name":937,"slug":938,"type":16},{"name":21,"slug":22,"type":16},{"name":954,"slug":955,"type":16},"Orchestration","orchestration","2026-07-25T05:56:34.255071",{"slug":958,"name":958,"fn":959,"description":960,"org":961,"tags":962,"stars":26,"repoUrl":27,"updatedAt":970},"cao-learning","report task outcomes and distill lessons","Report task outcomes and distill lessons so the team improves across runs — report_outcome after each unit of work, retrospector handoffs at natural boundaries, and applying injected lessons. Use in workflows that run repeatedly over similar work items. Requires memory.learning_enabled; degrade silently when the tools report disabled.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[963,966,969],{"name":964,"slug":965,"type":16},"Best Practices","best-practices",{"name":967,"slug":968,"type":16},"Engineering","engineering",{"name":14,"slug":15,"type":16},"2026-07-29T06:00:28.147989",{"slug":972,"name":972,"fn":973,"description":974,"org":975,"tags":976,"stars":26,"repoUrl":27,"updatedAt":983},"cao-mcp-apps","operate CAO MCP application surfaces","Enable, operate, and extend CAO's MCP Apps surface — the host-rendered fleet dashboard visible inside MCP App hosts (Claude Desktop, ChatGPT, VS Code Copilot, Goose, Postman). Use when the user says \"enable MCP Apps in CAO\", \"the ui:\u002F\u002Fcao views aren't rendering\", \"rebuild MCP Apps bundles\", \"add a new ui:\u002F\u002Fcao\u002F* view\", or \"configure the MCP Apps OAuth scope layer\". Operates on the CAO_MCP_APPS_ENABLED surface and cao_mcp_apps\u002F build system. Not for the localhost:9889 browser dashboard, not for plugins, providers, or session management.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[977,978,981,982],{"name":937,"slug":938,"type":16},{"name":979,"slug":980,"type":16},"CLI","cli",{"name":921,"slug":922,"type":16},{"name":927,"slug":928,"type":16},"2026-07-22T05:35:52.952289",{"slug":985,"name":985,"fn":986,"description":987,"org":988,"tags":989,"stars":26,"repoUrl":27,"updatedAt":997},"cao-memory","manage durable agent memory and preferences","Store, recall, and forget durable facts with CAO memory — user preferences, project conventions, decisions, and corrections that should persist across sessions and agents. Use proactively to check memory before asking the user, and to save anything worth remembering. Distinct from any provider-native memory.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[990,991,994],{"name":937,"slug":938,"type":16},{"name":992,"slug":993,"type":16},"Memory","memory",{"name":995,"slug":996,"type":16},"Productivity","productivity","2026-07-12T08:37:03.180421",{"slug":999,"name":999,"fn":1000,"description":1001,"org":1002,"tags":1003,"stars":26,"repoUrl":27,"updatedAt":1008},"cao-plugin","scaffold CAO agent plugins","Create a new CAO (CLI Agent Orchestrator) plugin. Use this skill whenever the user wants to add a plugin that reacts to CAO lifecycle or messaging events, scaffold a plugin package, understand plugin requirements, or integrate an external system (Discord, Slack, dashboards, logging, metrics) with CAO. Also use when the user asks what plugin events are available, how plugin discovery works, or how to install a plugin into a CAO environment.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1004,1005,1006,1007],{"name":937,"slug":938,"type":16},{"name":885,"slug":886,"type":16},{"name":979,"slug":980,"type":16},{"name":924,"slug":925,"type":16},"2026-07-12T08:36:49.784639",17]