[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CSH372F004Y9D7PJTVTF34":3,"skill-same:01M7CSH372F004Y9D7PJTVTF34":48},{"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":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7CSH372F004Y9D7PJTVTF34","release-playbook",null,"Use when you need the end-to-end merge -> deploy -> watch -> verdict sequence for a delivery mode, or a tool result you did not expect (refusal, park, tag exists)","makifbaysal\u002Ftasktrooper","makifbaysal","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F25031940?v=4","01M7CRG3ZCSYZSSEVDMW3DEYC8","Mehmet Akif Baysal","01M7CRG401FJ8YAXH5ER730XYV",122,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Frelease-engineer\u002Fskills\u002Frelease-playbook\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: release-playbook\ncategory: release\ndescription: Use when you need the end-to-end merge -> deploy -> watch -> verdict sequence for a delivery mode, or a tool result you did not expect (refusal, park, tag exists)\n---\n\n# Release playbook\n\nYou own a task from the moment the board signs it off in `done` until it either lands in `released` or bounces back to `need_revision` through a rollback. Four steps, in order, and each is only worth doing because the one before it happened: **merge → deploy → watch → finish or roll back.**\n\n## 1. Merge\n\n```\nmerge_task_pull_request({ task_id })\n```\n\nreturns, among the usual merge fields, `release`:\n\n```json\n{\n  \"release\": {\n    \"mode\": \"dispatch\",\n    \"release_id\": \"b1e2...\",\n    \"status\": \"pending\",\n    \"next\": \"call deploy_release, then watch_release\"\n  }\n}\n```\n\n`release.next` already tells you the next call — read it, do not guess. `release.unconfirmed: true` is a flag on a task merged some other way, not a fifth mode — stop, the card already says why. The modes:\n\n| `release.mode` | What it means | What you do |\n|---|---|---|\n| `none` | Nothing deploys this component (a library, docs) — the merge itself was the release. | Nothing. |\n| `batch` | This component (desktop, mobile and friends) ships in a human-cut batch: the merge joined the component's draft release. | Nothing — the task stays in `done` until a human cuts it. |\n| `on_merge` | The merge itself deploys (a workflow on push, or a provider's git integration). | `watch_release`. |\n| `dispatch` | A deploy workflow needs to be dispatched at the release tag. | `deploy_release`, then `watch_release`. |\n\nA merge refusal is final, not a retry — see the refusal table in your `done` column.\n\n## 2. Deploy\n\n```\ndeploy_release({ task_id })\n```\n\n**`dispatch` mode** dispatches the profile's workflow at the release tag (`release\u002F\u003Csha12>`) and moves the release to `deploying`. Then call `watch_release`.\n\n**A cut `batch` release** (status `pending`, mode `batch` — you are woken with it, you never cut it) does something different per executor:\n\n| Executor | What `deploy_release` does |\n|---|---|\n| `github_actions` | Creates the release tag at the cut commit; the repository's own tag-triggered workflow builds and publishes. A tag that already exists is a **real refusal** here — unlike `dispatch`, a batch version is never re-used (SemVer §3: a released version's contents never change). Comment \"tag vX.Y.Z already exists — re-cut this release with a new version on the Deploy tab\" instead of retrying. |\n| `local` | Runs the profile's command on this machine, in a detached worktree of the cut commit, and logs it (`local_run` on the release). |\n| `store` | Starts a store build for every platform the repository has a linked app for (`store_builds` on the release). |\n\nThen call `watch_release` the same as for `dispatch`.\n\n## 3. Watch\n\n```\nwatch_release({ task_id })\n```\n\nWhile the deploy is running or the soak window is open, this call **parks your card** and your run ends — a system sweeper is watching the deploy and the health\u002Fsmoke\u002Ferror window for you. Do not poll, sleep, or call it again in a loop; you are woken automatically when there is something to decide. When you are woken, `watch_release` (or `get_release`) returns the settled release instead of a park.\n\n## 4. Verdict — `awaiting_verdict`\n\nNever skip straight to a verdict because the deploy looked green — see `post-deploy-verification` for the exact calls, the verdict table, and the schema\u002Fflag\u002Fmulti-task edge cases. In short:\n\n```\nget_release({ task_id })                                   # checks already gathered, plus component\nquery_runtime_logs({ component, since: deployed_at })      # where the component has a bound runtime environment\nlist_runtime_errors({ component, since: \"\u003Cwell before deployed_at>\" })\n```\n\nClean evidence:\n\n```\nfinish_release({ task_id, note: \"logs clean since deployed_at, no new error groups (first_seen checked), smoke checks green\" })\n```\n\nEvidence of a problem:\n\n```\nrollback_release({ task_id, reason: \"verify_failed\", note: \"checkout smoke check returned 500 since the deploy; 3 error groups with first_seen after deployed_at, all in the checkout module\" })\n```\n\n`reason` is one of `deploy_failed` (the deploy itself never went green), `verify_failed` (it went green but the soak found a problem), `health_incident` (an incident opened inside the window, handled from the `released` column). `note` is required — state what you actually checked, not just the conclusion.\n\n## 5. Rollback outcomes\n\n`rollback_release` either executes the rollback (`release.status: \"rolling_back\"`, with `rollback.manual_steps`) or — when `auto_rollback` is off for the component — writes the proposal as a comment and returns `proposed: true` with nothing executed. Both are correct, successful outcomes for their situation. See `rollback-runbook` for schema safety, working through `manual_steps`, and what to do when the call refuses outright (too old, a newer release shipped, a newer one open).\n\n## 6. Failed deploys\n\nA release that lands in `failed` (the deploy itself never went green) is read the same way: `get_release`, then per executor — `github_actions` → `get_deploy_logs` if there is a failed job; a batch `local` run → `local_run.tail` (and its log path) on the release itself; a batch `store` build → `store_builds` names the platform and its error. Then `rollback_release` with reason `deploy_failed` if the bad code is live or sitting on the default branch — otherwise there is nothing to undo, report what failed and stop.\n\n## Tool results you will see\n\n| Result | Means | Do |\n|---|---|---|\n| `watch_release` returns a park (`ResourceBlock`) | Nothing further to do in this run. | End the run; the sweeper re-dispatches you. |\n| `deploy_release` refused (before-deploy steps pending, deploy dependency unreleased, unconfirmed profile) | Same refusal family as the merge. | Stop, do not retry; you are woken when the condition clears. |\n| Batch tag already exists | A version is never re-used. | Comment asking for a re-cut with a new version; never delete or move the existing tag. |\n| `finish_release` refused | The release is not `awaiting_verdict` any more — something else already moved it. | `get_release` and act on its current `next`. |\n\n## Multi-task releases\n\nA release can carry several tasks (a cut batch, or a release that superseded an earlier one) — `finish_release` releases all of them at once. Your verdict evidence covers every task in `release.tasks`, not only the card you were woken on; see `post-deploy-verification` §1.\n","# Release playbook\n\nYou own a task from the moment the board signs it off in `done` until it either lands in `released` or bounces back to `need_revision` through a rollback. Four steps, in order, and each is only worth doing because the one before it happened: **merge → deploy → watch → finish or roll back.**\n\n## 1. Merge\n\n```\nmerge_task_pull_request({ task_id })\n```\n\nreturns, among the usual merge fields, `release`:\n\n```json\n{\n  \"release\": {\n    \"mode\": \"dispatch\",\n    \"release_id\": \"b1e2...\",\n    \"status\": \"pending\",\n    \"next\": \"call deploy_release, then watch_release\"\n  }\n}\n```\n\n`release.next` already tells you the next call — read it, do not guess. `release.unconfirmed: true` is a flag on a task merged some other way, not a fifth mode — stop, the card already says why. The modes:\n\n| `release.mode` | What it means | What you do |\n|---|---|---|\n| `none` | Nothing deploys this component (a library, docs) — the merge itself was the release. | Nothing. |\n| `batch` | This component (desktop, mobile and friends) ships in a human-cut batch: the merge joined the component's draft release. | Nothing — the task stays in `done` until a human cuts it. |\n| `on_merge` | The merge itself deploys (a workflow on push, or a provider's git integration). | `watch_release`. |\n| `dispatch` | A deploy workflow needs to be dispatched at the release tag. | `deploy_release`, then `watch_release`. |\n\nA merge refusal is final, not a retry — see the refusal table in your `done` column.\n\n## 2. Deploy\n\n```\ndeploy_release({ task_id })\n```\n\n**`dispatch` mode** dispatches the profile's workflow at the release tag (`release\u002F\u003Csha12>`) and moves the release to `deploying`. Then call `watch_release`.\n\n**A cut `batch` release** (status `pending`, mode `batch` — you are woken with it, you never cut it) does something different per executor:\n\n| Executor | What `deploy_release` does |\n|---|---|\n| `github_actions` | Creates the release tag at the cut commit; the repository's own tag-triggered workflow builds and publishes. A tag that already exists is a **real refusal** here — unlike `dispatch`, a batch version is never re-used (SemVer §3: a released version's contents never change). Comment \"tag vX.Y.Z already exists — re-cut this release with a new version on the Deploy tab\" instead of retrying. |\n| `local` | Runs the profile's command on this machine, in a detached worktree of the cut commit, and logs it (`local_run` on the release). |\n| `store` | Starts a store build for every platform the repository has a linked app for (`store_builds` on the release). |\n\nThen call `watch_release` the same as for `dispatch`.\n\n## 3. Watch\n\n```\nwatch_release({ task_id })\n```\n\nWhile the deploy is running or the soak window is open, this call **parks your card** and your run ends — a system sweeper is watching the deploy and the health\u002Fsmoke\u002Ferror window for you. Do not poll, sleep, or call it again in a loop; you are woken automatically when there is something to decide. When you are woken, `watch_release` (or `get_release`) returns the settled release instead of a park.\n\n## 4. Verdict — `awaiting_verdict`\n\nNever skip straight to a verdict because the deploy looked green — see `post-deploy-verification` for the exact calls, the verdict table, and the schema\u002Fflag\u002Fmulti-task edge cases. In short:\n\n```\nget_release({ task_id })                                   # checks already gathered, plus component\nquery_runtime_logs({ component, since: deployed_at })      # where the component has a bound runtime environment\nlist_runtime_errors({ component, since: \"\u003Cwell before deployed_at>\" })\n```\n\nClean evidence:\n\n```\nfinish_release({ task_id, note: \"logs clean since deployed_at, no new error groups (first_seen checked), smoke checks green\" })\n```\n\nEvidence of a problem:\n\n```\nrollback_release({ task_id, reason: \"verify_failed\", note: \"checkout smoke check returned 500 since the deploy; 3 error groups with first_seen after deployed_at, all in the checkout module\" })\n```\n\n`reason` is one of `deploy_failed` (the deploy itself never went green), `verify_failed` (it went green but the soak found a problem), `health_incident` (an incident opened inside the window, handled from the `released` column). `note` is required — state what you actually checked, not just the conclusion.\n\n## 5. Rollback outcomes\n\n`rollback_release` either executes the rollback (`release.status: \"rolling_back\"`, with `rollback.manual_steps`) or — when `auto_rollback` is off for the component — writes the proposal as a comment and returns `proposed: true` with nothing executed. Both are correct, successful outcomes for their situation. See `rollback-runbook` for schema safety, working through `manual_steps`, and what to do when the call refuses outright (too old, a newer release shipped, a newer one open).\n\n## 6. Failed deploys\n\nA release that lands in `failed` (the deploy itself never went green) is read the same way: `get_release`, then per executor — `github_actions` → `get_deploy_logs` if there is a failed job; a batch `local` run → `local_run.tail` (and its log path) on the release itself; a batch `store` build → `store_builds` names the platform and its error. Then `rollback_release` with reason `deploy_failed` if the bad code is live or sitting on the default branch — otherwise there is nothing to undo, report what failed and stop.\n\n## Tool results you will see\n\n| Result | Means | Do |\n|---|---|---|\n| `watch_release` returns a park (`ResourceBlock`) | Nothing further to do in this run. | End the run; the sweeper re-dispatches you. |\n| `deploy_release` refused (before-deploy steps pending, deploy dependency unreleased, unconfirmed profile) | Same refusal family as the merge. | Stop, do not retry; you are woken when the condition clears. |\n| Batch tag already exists | A version is never re-used. | Comment asking for a re-cut with a new version; never delete or move the existing tag. |\n| `finish_release` refused | The release is not `awaiting_verdict` any more — something else already moved it. | `get_release` and act on its current `next`. |\n\n## Multi-task releases\n\nA release can carry several tasks (a cut batch, or a release that superseded an earlier one) — `finish_release` releases all of them at once. Your verdict evidence covers every task in `release.tasks`, not only the card you were woken on; see `post-deploy-verification` §1.\n",{"name":5,"category":24,"description":7},"release","main","910394a0b6db9fc8c9a8f36328ec421d3ebd8b38","catalog\u002Fagents\u002Frelease-engineer\u002Fskills\u002Frelease-playbook",18,"https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper","https:\u002F\u002Ftasktrooper.ai","https:\u002F\u002Fgithub.com\u002Fmakifbaysal","Senior Software Developer","2026-11-15 04:50:29",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add makifbaysal\u002Ftasktrooper --skill release-playbook",{"tool":40,"command":41},"bunx","bunx skills add makifbaysal\u002Ftasktrooper --skill release-playbook",{"tool":43,"command":44},"pnpm","pnpm dlx skills add makifbaysal\u002Ftasktrooper --skill release-playbook","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Frelease-engineer\u002Fskills\u002Frelease-playbook\u002FSKILL.md",true,[],[49,55,62,69,76,83],{"skillNo":50,"name":51,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M7CSVEPQW6J6TD6CS2T5FPNZ","visual-craft","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fvisual-craft\u002FSKILL.md",[],[],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":19},"01M7CSVAQG91ETHQ0HBNZDG5FR","ux-copy","Use when you write any words a user reads in a mockup or a hand-off - content language and voice, buttons, labels, errors, empty states, confirmations, numbers and dates, and the longest-string check","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fux-copy\u002FSKILL.md",[],[],{"skillNo":63,"name":64,"nameZh":6,"description":65,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":66,"categories":67,"keywords":68,"publishedAt":19},"01M7CSV6H0J38659NDXHJQ8396","screen-mockup-html","Use when you draw screens as the design task's HTML mockup - ONE canvas document holding every screen and every variant side by side, the sanitizer-safe recipe, phone and wide frames, the states grid, static control states, and the loopback self-check with screenshots at 375, 768 and 1440 attached to the task","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fscreen-mockup-html\u002FSKILL.md",[],[],{"skillNo":70,"name":71,"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},"01M7CSV3JFRF3BM3XK52M7YEYR","performance-awareness","How the performance score moves for your role. Use when the score context shows a declining trend, or before handing off work you have not self-checked against every criterion.","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fperformance-awareness\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M7CSTZDH00YRF5SCT7BRT9QD","design-variants","Use when a screen design needs variants - how many, variants as side-by-side rows on the one canvas, what makes two variants genuinely different, recommending one, how the human's choice is read, and narrowing to the chosen one in revision","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-variants\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M7CSTV6P3SM3PAE28DB0RF77","design-system-authoring","Use when you derive, change or extend a design system - where tokens live per stack, the DTCG token tree, role-based names for both themes, base vs layer vs drift, the DESIGN.md section order, the inventory format, the contrast check and the lint every proposal returns","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fdesign-system-authoring\u002FSKILL.md",[],[]]