[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M786XBTY8S47AMKWEVM905VV":3,"skill-same:01M786XBTY8S47AMKWEVM905VV":62},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":28,"commitSha":29,"skillDirectory":30,"forksCount":14,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":33,"install":34,"resources":47},"01M786XBTY8S47AMKWEVM905VV","subagent-driven-development",null,"Use when executing implementation plans, working on independent tasks, or when asked to dispatch subagents. Routes through sp-* workflows and dispatches native subagents from validated execution packets.","chenziyang110\u002Flaunchdeck","chenziyang110","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F43693913?v=4","01M7863EBWK1D0P6A29YGEMAS4","Robin11","01M7863ECM1VB6RXDMHTW90D2T",0,"MIT","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fsubagent-driven-development\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: \"subagent-driven-development\"\ndescription: \"Use when executing implementation plans, working on independent tasks, or when asked to dispatch subagents. Routes through sp-* workflows and dispatches native subagents from validated execution packets.\"\ncompatibility: \"Requires spec-kit project structure with .specify\u002F directory\"\nmetadata:\n  author: \"github-spec-kit\"\n  source: \"templates\u002Fpassive-skills\u002Fsubagent-driven-development\u002FSKILL.md\"\n---\n\n\n# Subagent-Driven Development (Spec Kit Plus)\n\nSpec Kit Plus keeps workflow state in `sp-*` artifacts, but it should still use\nnative subagents when bounded delegation materially improves throughput, isolation,\nspecialist quality, or verification confidence. Subagents are not a competing workflow; they are\nthe execution workers behind packetized `sp-quick` work after understanding\nconfirmation, broad or independent `sp-debug` evidence lanes, `sp-map-scan`,\n`sp-map-build`, and `sp-implement`.\n\n## Workflow-Owned Validation\n\nWhen `sp-implement` uses feature validation epochs, packets carry cheap\n`task_checks`, accepted baseline refs when needed, and test impact. The Leader\nowns every heavyweight validation epoch shared across Implement and Review.\nWorkers and this passive skill must not start an extra validation epoch for a\nTxx, dispatch, join, resume, or completion handoff.\n\n## Core Rule\n\nRoute first, choose the lightest safe execution surface, then packetize only selected delegated work.\n\n- Route into the smallest correct `sp-*` workflow before implementation or\n  investigation begins.\n- Use leader-direct execution for a small or tightly coupled task when delegation\n  adds no critical-path or quality benefit and the owning workflow permits it.\n- Compile and validate a `WorkerTaskPacket` just in time before selected delegated work.\n- Dispatch `one-subagent` when one safe lane is ready.\n- Dispatch `parallel-subagents` when two or more independent lanes can run\n  concurrently.\n- If dispatch fails, record the event and re-evaluate route safety. Execute locally\n  only when the task independently qualifies for leader-direct.\n- `sp-debug` may stay leader-inline for small focused investigations; use\n  subagent-assisted execution when the investigation exposes broad, independent,\n  or parallel evidence lanes.\n- Do not use old strategy labels as routing choices.\n- `sp-teams` only when Codex work needs durable team state, explicit join-point\n  tracking, result files, or lifecycle control beyond one in-session subagent\n  burst.\n\n## Process\n\n1. **Select the owning workflow**: Route to canonical `sp-tasks` when a plan lacks\n   an executable task graph. Route to canonical `sp-implement` for planned feature\n   execution, `sp-quick` for lightweight tracked work, `sp-debug` for root-cause\n   work, and the relevant map\u002Ftest workflow for project cognition or testing-system\n   lanes. When telling the user what to type, use `$sp-tasks`,\n   `$sp-implement`, `$sp-quick`, or `$sp-debug`.\n2. **Select the route**: Keep small\u002Ftightly coupled work leader-direct. For an\n   independent or parallel lane, compile a validated `WorkerTaskPacket` from the\n   current task, stable refs, and live code.\n3. **Dispatch in the current runtime**: Use the `native-subagents` surface such\n   as Codex `spawn_agent`, Claude Task, or the active CLI's equivalent. The\n   leader owns packet quality, lane selection, and integration, but should not\n   implement the lane locally while subagent execution is active. Use\n   `managed-team` only for durable team state or lifecycle control, and use\n   `leader-inline` only as the owning workflow's selected mode.\n4. **Join on evidence**: Wait for every subagent's structured handoff. The\n   handoff must name changed files, cheap task checks, test impact, accepted\n   epoch refs, failures, open risks, and any spec or plan gaps. An idle or silent\n   subagent is not completed work.\n5. **Review on triggers**: Run the single task reviewer only for drift, parallel\n   joins, scope violations, validation failure, worker concerns, obligation\n   conflicts, real-entrypoint gaps, or review-window triggers. Integrate verdict,\n   validation, and recovery into the task lifecycle record.\n\n## Dispatch Prompt Contract\n\nA subagent prompt must include:\n\n- The owning `sp-*` workflow and lane identifier.\n- The validated `WorkerTaskPacket` or equivalent packet summary.\n- Exact write set and paths the worker must not touch.\n- Stable authoritative refs from the current task, plan\u002Fspec contracts, live code,\n  and memory rules actually needed by the lane.\n- Accepted change-set RED\u002Fbaseline epoch ref or test-authoring-only status,\n  cheap task checks, and test impact for Leader-owned convergence.\n- For scan, build, PRD scan, and map-update evidence lanes, include explicit `assigned_paths` or changed paths. A subagent must not silently narrow assigned scope; if the set does not fit, the worker result returns top-level `acceptance=fail_gap`, marks affected paths with path-level `coverage[].outcome=\"overflow\"` or `coverage[].outcome=\"blocked\"`, and includes the smallest safe split or recovery suggestion. `overflow` and `blocked` are path or queue states, not top-level worker result acceptance values.\n- Worker results for mutable work must include changed paths, behavior surfaces, generated surfaces, state contracts, verification, known unknowns, and confidence notes so the parent workflow can build the inline project cognition update payload.\n- Required structured handoff format.\n\nThe leader must not dispatch from raw task text alone. If the packet is missing,\nambiguous, or does not describe boundaries and verification, create or repair the\npacket before dispatch.\n\n## Routing Rules\n\n- If the user says \"execute the plan\", \"start building\", \"implement tasks\", or\n  equivalent, route to `sp-implement`, or to `sp-tasks` first if tasks are not\n  generated. If a manual next step is needed, tell the user to run\n  `$sp-implement`, or `$sp-tasks` when the task graph is missing.\n- If the user asks to \"dispatch agents\", \"split this up\", or \"run subagents\",\n  route to the workflow that owns the state, then dispatch native subagents from\n  validated packets.\n- If the current workflow finds 2+ independent lanes, pair this skill with\n  `dispatching-parallel-agents`.\n- If the task is truly trivial and tightly coupled, `sp-fast` may stay inline.\n  Do not use \"small\" as a reason to skip routing for non-trivial work.\n\n## Red Flags\n\n- Treating every implementation task as mandatory delegation even when a small,\n  tightly coupled leader-direct route is safer and faster.\n- Doing leader-inline work after a `sp-debug` route selected\n  `subagent-assisted` execution or after independent evidence lanes are\n  available, without recording why the investigation remains small and focused.\n- Dispatching raw task text without a validated `WorkerTaskPacket`.\n- Asking the user to open separate terminals when native subagents are available\n  in the current runtime.\n- Treating an idle subagent as done work.\n- Accepting a handoff that lacks changed-file summary, cheap task-check evidence,\n  test impact, or its required epoch ref.\n- Skipping event-triggered review when the workflow records a review trigger.\n- Updating `tasks.md`, quick status, or workflow state as complete before the\n  structured handoff and validation evidence exist.\n","# Subagent-Driven Development (Spec Kit Plus)\n\nSpec Kit Plus keeps workflow state in `sp-*` artifacts, but it should still use\nnative subagents when bounded delegation materially improves throughput, isolation,\nspecialist quality, or verification confidence. Subagents are not a competing workflow; they are\nthe execution workers behind packetized `sp-quick` work after understanding\nconfirmation, broad or independent `sp-debug` evidence lanes, `sp-map-scan`,\n`sp-map-build`, and `sp-implement`.\n\n## Workflow-Owned Validation\n\nWhen `sp-implement` uses feature validation epochs, packets carry cheap\n`task_checks`, accepted baseline refs when needed, and test impact. The Leader\nowns every heavyweight validation epoch shared across Implement and Review.\nWorkers and this passive skill must not start an extra validation epoch for a\nTxx, dispatch, join, resume, or completion handoff.\n\n## Core Rule\n\nRoute first, choose the lightest safe execution surface, then packetize only selected delegated work.\n\n- Route into the smallest correct `sp-*` workflow before implementation or\n  investigation begins.\n- Use leader-direct execution for a small or tightly coupled task when delegation\n  adds no critical-path or quality benefit and the owning workflow permits it.\n- Compile and validate a `WorkerTaskPacket` just in time before selected delegated work.\n- Dispatch `one-subagent` when one safe lane is ready.\n- Dispatch `parallel-subagents` when two or more independent lanes can run\n  concurrently.\n- If dispatch fails, record the event and re-evaluate route safety. Execute locally\n  only when the task independently qualifies for leader-direct.\n- `sp-debug` may stay leader-inline for small focused investigations; use\n  subagent-assisted execution when the investigation exposes broad, independent,\n  or parallel evidence lanes.\n- Do not use old strategy labels as routing choices.\n- `sp-teams` only when Codex work needs durable team state, explicit join-point\n  tracking, result files, or lifecycle control beyond one in-session subagent\n  burst.\n\n## Process\n\n1. **Select the owning workflow**: Route to canonical `sp-tasks` when a plan lacks\n   an executable task graph. Route to canonical `sp-implement` for planned feature\n   execution, `sp-quick` for lightweight tracked work, `sp-debug` for root-cause\n   work, and the relevant map\u002Ftest workflow for project cognition or testing-system\n   lanes. When telling the user what to type, use `$sp-tasks`,\n   `$sp-implement`, `$sp-quick`, or `$sp-debug`.\n2. **Select the route**: Keep small\u002Ftightly coupled work leader-direct. For an\n   independent or parallel lane, compile a validated `WorkerTaskPacket` from the\n   current task, stable refs, and live code.\n3. **Dispatch in the current runtime**: Use the `native-subagents` surface such\n   as Codex `spawn_agent`, Claude Task, or the active CLI's equivalent. The\n   leader owns packet quality, lane selection, and integration, but should not\n   implement the lane locally while subagent execution is active. Use\n   `managed-team` only for durable team state or lifecycle control, and use\n   `leader-inline` only as the owning workflow's selected mode.\n4. **Join on evidence**: Wait for every subagent's structured handoff. The\n   handoff must name changed files, cheap task checks, test impact, accepted\n   epoch refs, failures, open risks, and any spec or plan gaps. An idle or silent\n   subagent is not completed work.\n5. **Review on triggers**: Run the single task reviewer only for drift, parallel\n   joins, scope violations, validation failure, worker concerns, obligation\n   conflicts, real-entrypoint gaps, or review-window triggers. Integrate verdict,\n   validation, and recovery into the task lifecycle record.\n\n## Dispatch Prompt Contract\n\nA subagent prompt must include:\n\n- The owning `sp-*` workflow and lane identifier.\n- The validated `WorkerTaskPacket` or equivalent packet summary.\n- Exact write set and paths the worker must not touch.\n- Stable authoritative refs from the current task, plan\u002Fspec contracts, live code,\n  and memory rules actually needed by the lane.\n- Accepted change-set RED\u002Fbaseline epoch ref or test-authoring-only status,\n  cheap task checks, and test impact for Leader-owned convergence.\n- For scan, build, PRD scan, and map-update evidence lanes, include explicit `assigned_paths` or changed paths. A subagent must not silently narrow assigned scope; if the set does not fit, the worker result returns top-level `acceptance=fail_gap`, marks affected paths with path-level `coverage[].outcome=\"overflow\"` or `coverage[].outcome=\"blocked\"`, and includes the smallest safe split or recovery suggestion. `overflow` and `blocked` are path or queue states, not top-level worker result acceptance values.\n- Worker results for mutable work must include changed paths, behavior surfaces, generated surfaces, state contracts, verification, known unknowns, and confidence notes so the parent workflow can build the inline project cognition update payload.\n- Required structured handoff format.\n\nThe leader must not dispatch from raw task text alone. If the packet is missing,\nambiguous, or does not describe boundaries and verification, create or repair the\npacket before dispatch.\n\n## Routing Rules\n\n- If the user says \"execute the plan\", \"start building\", \"implement tasks\", or\n  equivalent, route to `sp-implement`, or to `sp-tasks` first if tasks are not\n  generated. If a manual next step is needed, tell the user to run\n  `$sp-implement`, or `$sp-tasks` when the task graph is missing.\n- If the user asks to \"dispatch agents\", \"split this up\", or \"run subagents\",\n  route to the workflow that owns the state, then dispatch native subagents from\n  validated packets.\n- If the current workflow finds 2+ independent lanes, pair this skill with\n  `dispatching-parallel-agents`.\n- If the task is truly trivial and tightly coupled, `sp-fast` may stay inline.\n  Do not use \"small\" as a reason to skip routing for non-trivial work.\n\n## Red Flags\n\n- Treating every implementation task as mandatory delegation even when a small,\n  tightly coupled leader-direct route is safer and faster.\n- Doing leader-inline work after a `sp-debug` route selected\n  `subagent-assisted` execution or after independent evidence lanes are\n  available, without recording why the investigation remains small and focused.\n- Dispatching raw task text without a validated `WorkerTaskPacket`.\n- Asking the user to open separate terminals when native subagents are available\n  in the current runtime.\n- Treating an idle subagent as done work.\n- Accepting a handoff that lacks changed-file summary, cheap task-check evidence,\n  test impact, or its required epoch ref.\n- Skipping event-triggered review when the workflow records a review trigger.\n- Updating `tasks.md`, quick status, or workflow state as complete before the\n  structured handoff and validation evidence exist.\n",{"name":5,"metadata":24,"description":7,"compatibility":27},{"author":25,"source":26},"github-spec-kit","templates\u002Fpassive-skills\u002Fsubagent-driven-development\u002FSKILL.md","Requires spec-kit project structure with .specify\u002F directory","main","021f5cbe02157229460a577e9a585f2da6d78cb0",".codex\u002Fskills\u002Fsubagent-driven-development","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck","https:\u002F\u002Fgithub.com\u002Fchenziyang110","2026-08-07 18:10:54",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add chenziyang110\u002Flaunchdeck --skill subagent-driven-development",{"tool":40,"command":41},"bunx","bunx skills add chenziyang110\u002Flaunchdeck --skill subagent-driven-development",{"tool":43,"command":44},"pnpm","pnpm dlx skills add chenziyang110\u002Flaunchdeck --skill subagent-driven-development","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fsubagent-driven-development\u002FSKILL.md",true,[48,54,58],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"code-quality-reviewer-prompt.md","REFERENCE","参考文档",1379,"https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fsubagent-driven-development\u002Fcode-quality-reviewer-prompt.md",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"implementer-prompt.md",5209,"https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fsubagent-driven-development\u002Fimplementer-prompt.md",{"path":59,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":60,"sourceHtmlUrl":61},"spec-reviewer-prompt.md",2295,"https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fsubagent-driven-development\u002Fspec-reviewer-prompt.md",[63,70,76,82,89,98],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7875F3HQ1GF5F6BR36KY3JE","launchdeck-agent","Codex host adapter contract fixture for Launchdeck.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fagent-hosts\u002Fcodex\u002Fskill\u002FSKILL.md",[],[],{"skillNo":71,"name":65,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":19},"01M7875DV2DQR33K1NFJHKAMZP","Claude host adapter contract fixture for Launchdeck.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fagent-hosts\u002Fclaude\u002Fskill\u002FSKILL.md",[],[],{"skillNo":77,"name":65,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M78759JQ7A6FQ8KDQTNYS3X9","Use this skill aggressively for local project lifecycle operations even when the user does not mention Launchdeck: start this project, run the dev server, dev, build, test, restart service, stop service, show logs, inspect ports, port is occupied, clean cache; and when the user explicitly asks to configure this local project for Launchdeck or create or generate a project-adapted `.launchdeck.yml`, including 为这个项目生成 Launchdeck 配置. Do not use for ordinary code edits, docs-only work, general explanations, API design discussion, or production deployment unless the user also asks to operate or configure a local project or service.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Fagent\u002Finstaller-payload\u002Fskill\u002Flaunchdeck-agent\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M78753ZJQ7G2CZT1W1G8WRJV","zig-testing","Use ONLY when user explicitly requests Zig test code, built-in test runner configuration,\ntest failure diagnosis, or coverage for Zig projects.\n\nTriggers: \"write test zig\", \"zig test\", \"testing.allocator\", \"zig coverage\",\n\"comptime test\", \"zig expectError\", \"zig snapshot test\", \"zig fuzz\".\n\nDo NOT activate for: general Zig coding without test context, build.zig configuration\nfor non-test targets, Zig compiler errors, or Zig build system setup.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fzig-testing\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":93,"licenseFileUrl":94,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M78750JXY4Z5D5W1EX3PQ31Z","webapp-testing","Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fwebapp-testing\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fwebapp-testing\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7874X2T83EQQFRCEY5SJJVC","verification-before-completion","Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fverification-before-completion\u002FSKILL.md",[],[]]