[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CSGS19RBPMR5C3Q8JBRY4G":3,"skill-same:01M7CSGS19RBPMR5C3Q8JBRY4G":49},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":31,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":33,"updatedAt":34,"install":35,"resources":48},"01M7CSGS19RBPMR5C3Q8JBRY4G","post-deploy-verification",null,"Use when a release reaches awaiting_verdict (or you are woken for an incident) - how to read runtime logs, error groups, health and smoke evidence for the release's component and decide finish, re-check, roll back or hand to a human","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\u002Fpost-deploy-verification\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: post-deploy-verification\ncategory: release\ndescription: Use when a release reaches awaiting_verdict (or you are woken for an incident) - how to read runtime logs, error groups, health and smoke evidence for the release's component and decide finish, re-check, roll back or hand to a human\nsource: obra\u002Fsuperpowers (MIT), adapted\n---\n\n# Post-deploy verification\n\nA deploy that went green tells you the process started. It says nothing about whether the change it shipped actually works, or whether it broke something else that happened to keep running. Verification is what closes that gap, and it happens every time — not only when something looks wrong.\n\n## 1. Procedure — exact calls\n\n1. `get_release({ task_id })` → read `deployed_at`, `component`, `checks` (health samples, smoke results, `new_errors`, `notes`, `early_stop`), `profile.auto_rollback`, and `tasks[]` — every task this release covers, not only the one you were woken on.\n2. `get_environment` only if you need to confirm nothing is bound (`checks.notes` already says so in most cases).\n3. `list_runtime_errors({ component, since: \"\u003Cwindow that starts well before deployed_at>\" })` — see \"Reading error groups\" below for the window and how to judge `new`.\n4. `query_runtime_logs({ component, since: deployed_at, min_severity: \"error\" })`; re-query with `text:` set to a route or module the release changed when the error list alone is not enough.\n5. `run_smoke_checks({ task_id })` when the stored evidence is older than ~30 minutes, or when `early_stop` itself came from a smoke failure and you need a fresh read before deciding.\n6. One read-only acceptance check per observable criterion of **every** task in `release.tasks` (`list_acceptance_criteria` per task) — a release with several tasks is verified in full, not just the card that woke you.\n\n## 2. Reading error groups\n\n`list_runtime_errors`' `new: true` is accurate on a provider with native grouping (gcloud). On a provider without one (Vercel, AWS — fallback fingerprint grouping), it only means \"this group's first occurrence fell in the back half of whatever window you queried\" — querying `since: deployed_at` puts a regression that starts right after the deploy in the window's *front* half, so it can read `new: false` and never show up as a concern.\n\n- Judge a group by its own `first_seen` against `release.deployed_at`, never by `new` alone.\n- Read with a window that starts well before `deployed_at` — roughly twice the time elapsed since the deploy, or at least the soak period — so an occurrence from before the deploy is visible and can be told apart from one that started with it.\n- Before rolling back on one low-count group, re-read with `since: \"1d\"`. If the same message recurs before `deployed_at`, it is chronic, not a regression. Once you have confirmed a group is chronic and unrelated to any release, `save_memory` (project scope) a one-line note of its fingerprint message (\"`redis timeout` in apps\u002Fapi recurs hourly; not a release signal\") so a future run of this skill recognizes it without a full 1-day re-read.\n- `checks.new_errors` (the sweeper's own field) can be empty on Vercel or AWS even when errors genuinely began with the deploy — your own read in this run is the check, not a formality that confirms what the sweeper already decided.\n\n## 3. Verdict table\n\nEvidence maps to one of four outcomes — finish, re-check once, roll back, or hand to a human — the same shape as Argo Rollouts' Successful\u002FFailed\u002FInconclusive and Flagger's \"N failed checks before rollback\": one blip is not a trend, and an unreadable signal is not a pass.\n\n| Evidence | Verdict |\n|---|---|\n| `early_stop: smoke check failed` | Re-run `run_smoke_checks` once. Passes, and no error group has `first_seen >= deployed_at` → finish, naming the transient failure and the re-check. Fails again → `rollback_release` (`verify_failed`). |\n| Health check failed twice | `fetch_url` the `checks.health_url` once (GET). Still failing → roll back. Recovered → look for a restart\u002Fcrash line in the logs since `deployed_at`; clean → finish, noting the blip. |\n| An error group with `first_seen >= deployed_at` whose message, path or stack touches a file or route this release's tasks changed (`get_task_pull_request({ task_id, include_diff: false })` → files) | Roll back (`verify_failed`). |\n| Same, but unrelated and it also recurs before the deploy in a `1d` read | Finish, naming each group and why it is not a regression. |\n| Only structural gaps in `notes` (no bound environment, no health URL, no base URL) | Finish, stating the gap verbatim — it is a different kind of verification, not a failed one. |\n| A runtime tool itself errored | Retry once. Still failing, and health and smoke are both green → finish, saying the logs were unreadable this run. |\n| **Nothing at all is readable** for an `on_merge`\u002F`dispatch` service (no health, no smoke, no logs, and the retry above also failed) | One comment: \"verdict needs a human: no evidence was readable (`\u003Cwhat you tried>`)\" — and stop. A human finishes or rolls back from the Deploy tab. This is the one case where \"finish with the gap stated\" is not the default: with nothing readable there is no evidence to finish *on*. |\n\n### Recording a missing deploy target\n\nA `notes` gap like \"no health URL\" or \"no base URL\" does not always mean nobody can fix it — if the deploy's own output (`get_deploy_logs`, a `local_run.tail`, or a workflow log) names the real URL the service answers on, `update_deploy_target({ env, base_url \u002F health_url \u002F logs_url })` records it so the *next* soak and verdict actually have something to read, instead of hitting the same gap every release. Still finish or roll back on the evidence you have in THIS run — recording the target fixes the next one, it does not retroactively create evidence for this one.\n\n## 4. Feature-flagged changes\n\nIf a task's `after_deploy` (on `release.tasks[]`) enables a flag or config, its new behaviour is not reachable yet — that is expected, not a failure. Check that the old path still works, and say \"criterion X is behind flag Y, enabled after release by a human\" in your note. A flag flip is a manual step for a human, never something you do yourself.\n\n## 5. Notes that count as evidence\n\n✅ \"query_runtime_logs (apps\u002Fapi, since 2026-10-01T10:02Z, ≥error): 0 entries; list_runtime_errors 40m: 2 groups, both first_seen before deploy (chronic `redis timeout`, 1d read shows it hourly); smoke 3\u002F3 (re-run 10:41); T-12 GET \u002Fapi\u002Forders?status=paid → 200 with `paid_at` field.\"\n\n❌ \"Looks fine, logs clean.\"\n\n## 6. Rationalizations (adapted from obra\u002Fsuperpowers)\n\n- \"Deploy went green\" → that is the process starting, not the behaviour it shipped.\n- \"Soak was green\" → read it yourself, in THIS run — the sweeper's own fields are a starting point, not a substitute.\n- \"`new_errors` is empty\" → on Vercel or AWS that can miss an error that genuinely started with the deploy; read `first_seen` yourself.\n- \"The error is probably old\" → show the `first_seen`, don't assert it.\n- \"One failed probe\" → re-check once, then decide — never roll back, and never finish, on a single sample.\n\n## Common Mistakes\n\n- Omitting `component` in a monorepo — silently reads the root component's logs, not the release's own.\n- Reading errors with the default 24h window instead of a window anchored before `deployed_at`.\n- Judging novelty by `new` instead of `first_seen` against `deployed_at`.\n- Verifying only the task you were woken on when the release carries several.\n- Writing to production (seeding, creating records) to check a criterion.\n- \"Should\", \"probably\", \"seems\" anywhere in a finish\u002Frollback note.\n\n## Batch releases with no runtime environment\n\nMost desktop and mobile components have no bound runtime environment at all — there is no host to sample health from, no log stream to query, no error groups to list. That is not an incomplete verification, it is a different one: see `batch-artifact-verification` for what to check instead (the published artifact, store build, or local run), and say explicitly in the `finish_release`\u002F`rollback_release` note that no runtime environment is bound — the same rule as any other gap: never let silence read as a pass.\n","# Post-deploy verification\n\nA deploy that went green tells you the process started. It says nothing about whether the change it shipped actually works, or whether it broke something else that happened to keep running. Verification is what closes that gap, and it happens every time — not only when something looks wrong.\n\n## 1. Procedure — exact calls\n\n1. `get_release({ task_id })` → read `deployed_at`, `component`, `checks` (health samples, smoke results, `new_errors`, `notes`, `early_stop`), `profile.auto_rollback`, and `tasks[]` — every task this release covers, not only the one you were woken on.\n2. `get_environment` only if you need to confirm nothing is bound (`checks.notes` already says so in most cases).\n3. `list_runtime_errors({ component, since: \"\u003Cwindow that starts well before deployed_at>\" })` — see \"Reading error groups\" below for the window and how to judge `new`.\n4. `query_runtime_logs({ component, since: deployed_at, min_severity: \"error\" })`; re-query with `text:` set to a route or module the release changed when the error list alone is not enough.\n5. `run_smoke_checks({ task_id })` when the stored evidence is older than ~30 minutes, or when `early_stop` itself came from a smoke failure and you need a fresh read before deciding.\n6. One read-only acceptance check per observable criterion of **every** task in `release.tasks` (`list_acceptance_criteria` per task) — a release with several tasks is verified in full, not just the card that woke you.\n\n## 2. Reading error groups\n\n`list_runtime_errors`' `new: true` is accurate on a provider with native grouping (gcloud). On a provider without one (Vercel, AWS — fallback fingerprint grouping), it only means \"this group's first occurrence fell in the back half of whatever window you queried\" — querying `since: deployed_at` puts a regression that starts right after the deploy in the window's *front* half, so it can read `new: false` and never show up as a concern.\n\n- Judge a group by its own `first_seen` against `release.deployed_at`, never by `new` alone.\n- Read with a window that starts well before `deployed_at` — roughly twice the time elapsed since the deploy, or at least the soak period — so an occurrence from before the deploy is visible and can be told apart from one that started with it.\n- Before rolling back on one low-count group, re-read with `since: \"1d\"`. If the same message recurs before `deployed_at`, it is chronic, not a regression. Once you have confirmed a group is chronic and unrelated to any release, `save_memory` (project scope) a one-line note of its fingerprint message (\"`redis timeout` in apps\u002Fapi recurs hourly; not a release signal\") so a future run of this skill recognizes it without a full 1-day re-read.\n- `checks.new_errors` (the sweeper's own field) can be empty on Vercel or AWS even when errors genuinely began with the deploy — your own read in this run is the check, not a formality that confirms what the sweeper already decided.\n\n## 3. Verdict table\n\nEvidence maps to one of four outcomes — finish, re-check once, roll back, or hand to a human — the same shape as Argo Rollouts' Successful\u002FFailed\u002FInconclusive and Flagger's \"N failed checks before rollback\": one blip is not a trend, and an unreadable signal is not a pass.\n\n| Evidence | Verdict |\n|---|---|\n| `early_stop: smoke check failed` | Re-run `run_smoke_checks` once. Passes, and no error group has `first_seen >= deployed_at` → finish, naming the transient failure and the re-check. Fails again → `rollback_release` (`verify_failed`). |\n| Health check failed twice | `fetch_url` the `checks.health_url` once (GET). Still failing → roll back. Recovered → look for a restart\u002Fcrash line in the logs since `deployed_at`; clean → finish, noting the blip. |\n| An error group with `first_seen >= deployed_at` whose message, path or stack touches a file or route this release's tasks changed (`get_task_pull_request({ task_id, include_diff: false })` → files) | Roll back (`verify_failed`). |\n| Same, but unrelated and it also recurs before the deploy in a `1d` read | Finish, naming each group and why it is not a regression. |\n| Only structural gaps in `notes` (no bound environment, no health URL, no base URL) | Finish, stating the gap verbatim — it is a different kind of verification, not a failed one. |\n| A runtime tool itself errored | Retry once. Still failing, and health and smoke are both green → finish, saying the logs were unreadable this run. |\n| **Nothing at all is readable** for an `on_merge`\u002F`dispatch` service (no health, no smoke, no logs, and the retry above also failed) | One comment: \"verdict needs a human: no evidence was readable (`\u003Cwhat you tried>`)\" — and stop. A human finishes or rolls back from the Deploy tab. This is the one case where \"finish with the gap stated\" is not the default: with nothing readable there is no evidence to finish *on*. |\n\n### Recording a missing deploy target\n\nA `notes` gap like \"no health URL\" or \"no base URL\" does not always mean nobody can fix it — if the deploy's own output (`get_deploy_logs`, a `local_run.tail`, or a workflow log) names the real URL the service answers on, `update_deploy_target({ env, base_url \u002F health_url \u002F logs_url })` records it so the *next* soak and verdict actually have something to read, instead of hitting the same gap every release. Still finish or roll back on the evidence you have in THIS run — recording the target fixes the next one, it does not retroactively create evidence for this one.\n\n## 4. Feature-flagged changes\n\nIf a task's `after_deploy` (on `release.tasks[]`) enables a flag or config, its new behaviour is not reachable yet — that is expected, not a failure. Check that the old path still works, and say \"criterion X is behind flag Y, enabled after release by a human\" in your note. A flag flip is a manual step for a human, never something you do yourself.\n\n## 5. Notes that count as evidence\n\n✅ \"query_runtime_logs (apps\u002Fapi, since 2026-10-01T10:02Z, ≥error): 0 entries; list_runtime_errors 40m: 2 groups, both first_seen before deploy (chronic `redis timeout`, 1d read shows it hourly); smoke 3\u002F3 (re-run 10:41); T-12 GET \u002Fapi\u002Forders?status=paid → 200 with `paid_at` field.\"\n\n❌ \"Looks fine, logs clean.\"\n\n## 6. Rationalizations (adapted from obra\u002Fsuperpowers)\n\n- \"Deploy went green\" → that is the process starting, not the behaviour it shipped.\n- \"Soak was green\" → read it yourself, in THIS run — the sweeper's own fields are a starting point, not a substitute.\n- \"`new_errors` is empty\" → on Vercel or AWS that can miss an error that genuinely started with the deploy; read `first_seen` yourself.\n- \"The error is probably old\" → show the `first_seen`, don't assert it.\n- \"One failed probe\" → re-check once, then decide — never roll back, and never finish, on a single sample.\n\n## Common Mistakes\n\n- Omitting `component` in a monorepo — silently reads the root component's logs, not the release's own.\n- Reading errors with the default 24h window instead of a window anchored before `deployed_at`.\n- Judging novelty by `new` instead of `first_seen` against `deployed_at`.\n- Verifying only the task you were woken on when the release carries several.\n- Writing to production (seeding, creating records) to check a criterion.\n- \"Should\", \"probably\", \"seems\" anywhere in a finish\u002Frollback note.\n\n## Batch releases with no runtime environment\n\nMost desktop and mobile components have no bound runtime environment at all — there is no host to sample health from, no log stream to query, no error groups to list. That is not an incomplete verification, it is a different one: see `batch-artifact-verification` for what to check instead (the published artifact, store build, or local run), and say explicitly in the `finish_release`\u002F`rollback_release` note that no runtime environment is bound — the same rule as any other gap: never let silence read as a pass.\n",{"name":5,"source":24,"category":25,"description":7},"obra\u002Fsuperpowers (MIT), adapted","release","main","910394a0b6db9fc8c9a8f36328ec421d3ebd8b38","catalog\u002Fagents\u002Frelease-engineer\u002Fskills\u002Fpost-deploy-verification",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":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add makifbaysal\u002Ftasktrooper --skill post-deploy-verification",{"tool":41,"command":42},"bunx","bunx skills add makifbaysal\u002Ftasktrooper --skill post-deploy-verification",{"tool":44,"command":45},"pnpm","pnpm dlx skills add makifbaysal\u002Ftasktrooper --skill post-deploy-verification","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Frelease-engineer\u002Fskills\u002Fpost-deploy-verification\u002FSKILL.md",true,[],[50,56,63,70,77,84],{"skillNo":51,"name":52,"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":53,"categories":54,"keywords":55,"publishedAt":19},"01M7CSVEPQW6J6TD6CS2T5FPNZ","visual-craft","https:\u002F\u002Fgithub.com\u002Fmakifbaysal\u002Ftasktrooper\u002Fblob\u002Fmain\u002Fcatalog\u002Fagents\u002Fui-designer\u002Fskills\u002Fvisual-craft\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"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":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},"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":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"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":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"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":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"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",[],[]]