[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7A0ASRWFJWRAMG6G33E837Q":3,"skill-same:01M7A0ASRWFJWRAMG6G33E837Q":45},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M7A0ASRWFJWRAMG6G33E837Q","long-running",null,"Detach and wait on any long worktree job (pack, notarize, full test, deploy) so a new chat turn does not SIGTERM it, and the human does not have to poll. Use when a command may run more than ~30s, or the user mentions wait \u002F status \u002F aborted builds. Do not use for a server that should stay up (a local site preview). Stop that job instead of waiting.","mattlevine\u002Fsideboard","mattlevine","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F836225?v=4","01M7A0A4TN8FJQD7Q6MB7K1HK1","Matt Levine","01M7A0A4V7C06D7JKTFA89TEM7",4,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Flong-running\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: long-running\ndescription: >-\n  Detach and wait on any long worktree job (pack, notarize, full test, deploy)\n  so a new chat turn does not SIGTERM it, and the human does not have to poll.\n  Use when a command may run more than ~30s, or the user mentions wait \u002F status\n  \u002F aborted builds. Do not use for a server that should stay up (a local site\n  preview). Stop that job instead of waiting.\n---\n\n# Long-running jobs\n\nSideboard ships this method to **every** worktree agent (`\u002Flong-running` is a product skill). This repo copy adds Mac-release notes below. A workspace skill with the same name wins over the bundled copy.\n\nA Sideboard or Cursor **worktree turn** SIGTERMs the agent shell (and its process group) when the user sends another message or the turn is interrupted. `block_until_ms: 0` is not enough — the child stays in that group.\n\nDo **not** ask the human to check back. Detach, then **wait** in 45s slices (same idea as MCP `wait_for_turn`) until `stillRunning` is false. Do not end the turn after starting a job — a native CLI Task or bash notify does **not** resume a Sideboard chat (Claude, Cursor, Codex, OpenCode).\n\nA process that is supposed to stay up is not this skill. Do not `start` a static file server, preview server, or anything else that never exits (`python -m http.server`, a dev server). `stillRunning` stays true, and the orchestrator will not let the turn end. To show the local marketing page, open `site\u002Findex.html` (side column or a browser). If a stay-up server is already a detached job, `stop` that id and continue. Do not loop `wait` on it.\n\n## Tool\n\nPrefer the helper path from the Sideboard playbook (absolute `node \"…\" start`). In this repo, `scripts\u002Fdetached-job.cjs` is the same file.\n\n```bash\n# Start (exits in ~1s; job survives this turn)\nnode scripts\u002Fdetached-job.cjs start \u003Cid> -- \u003Ccommand> [args...]\n\n# Wait — returns within ~45s even if the job is still going\nnode scripts\u002Fdetached-job.cjs wait \u003Cid>\n\n# Block until the process exits (humans \u002F a turn that will not be interrupted)\nnode scripts\u002Fdetached-job.cjs wait \u003Cid> --until-done\n\nnode scripts\u002Fdetached-job.cjs stop \u003Cid> [--reason TEXT]\nnode scripts\u002Fdetached-job.cjs status \u003Cid>\nnode scripts\u002Fdetached-job.cjs ui \u003Cid> [--title TEXT] [--out FILE]\n```\n\n`\u003Cid>` is kebab-case (`mac-release`, `core-test`, `fly-deploy`). State is `.context\u002F.sideboard\u002Fdetached-jobs\u002F\u003Cid>\u002F` (local scratch).\n\nWait JSON:\n\n- `stillRunning: true` → exit 2 → **call wait again**. Progress is in `progress` \u002F `phase`. Do not start a second job. Do not ping the user. If it is hanging, producing no useful output, or doing the wrong thing, call MCP `stop_job` (or `detached-job.cjs stop`) instead of looping forever.\n- `ok: true` → exit 0 → continue the rest of the task (README, commit, deploy next step).\n- `failed: true` → exit 1 → read `progress`, fix, start **once**.\n- `stopped: true` → you ended it. Present the last delta (`status=failed`) and decide the next step.\n\n## Sideboard UI (append-only log)\n\nThe side column **is** the live view. `wait_for_job` \u002F `stop_job` (and shell `detached-job` wait JSON) open a `type=log` pane automatically (`artifact_id` = job id, content = `delta`). You may still call `present_artifact` **`type=log`** for a custom title. Each update **appends** — send only `delta`, not the full log and not HTML.\n\n```\npresent_artifact\n  title: \u003Cshort title>\n  type: log\n  artifact_id: \u003Cid>\n  content: \u003Cwait.delta>          # new lines only; empty is ok\n  status: running | ok | failed  # from wait.status\n  phase: \u003Cwait.phase>\n```\n\nAfter **start**, the first `wait` opens the pane (`status=running`). Do **not** dump the log in chat. Do **not** also fence HTML.\n\n`detached-job.cjs ui` still writes a snapshot HTML file for debugging. Prefer `type=log`.\n\nIf you started via `release-mac-detached.js` (legacy pid\u002Flog):\n\n```bash\nnode scripts\u002Fdetached-job.cjs ui --pid-file apps\u002Fdesktop\u002Frelease\u002Frelease.pid --log-file apps\u002Fdesktop\u002Frelease\u002Frelease.log --ok-pattern RELEASE_BUILD_OK --title \"Mac pack\"\n```\n\nAd-hoc pid\u002Flog (legacy or another tool’s files):\n\n```bash\nnode scripts\u002Fdetached-job.cjs wait --pid-file FILE --log-file FILE [--ok-pattern TEXT]\n```\n\n## Agent loop\n\n1. `start` once. If JSON says `already-running`, do not start again.\n2. Loop `wait` (use `--timeout-ms 15000` for a livelier pane). The human should see **working** in the side column from the wait JSON — do not also dump the log in chat.\n3. After each slice, the same log pane appends `delta`. Optional: `present_artifact` the same id for a custom title.\n4. On `ok`, present once more (`status=ok`, last `delta`) and finish the task. On `failed`, fix from the log.\n5. **Stop** when the job is hanging, buffering forever, watching the wrong thing, or you already have the answer. MCP `stop_job` (same id, optional `reason`) or `node scripts\u002Fdetached-job.cjs stop \u003Cid>`. Do **not** stop a pack\u002Ftest\u002Fdeploy that is clearly making progress.\n\nNever tell the user “say status when it’s done.” You wait — unless you decided to stop.\n\n## Connector CLIs (Vercel, Supabase, Sentry, PostHog)\n\nHuge `--json` \u002F `--expand` dumps crash any worktree agent mid-turn (Claude \u002F Cursor \u002F Codex \u002F OpenCode). On Cursor the chat shows the packaged bundle (`file:\u002F\u002F…\u002Fcursor-runtime\u002F…\u002F@cursor\u002Fsdk\u002Fdist\u002Fesm\u002Findex.js` then minified `importas e from\"@bufbuild\u002Fprotobuf\"`). Write CLI\u002FHTTP output to `.context\u002Fcli\u002F` (not `.context\u002Fattachments\u002F`), then read a slice. Tight window first; one query at a time; `jq` only the fields you need. After the error line, stop fetching. If a raw shell is still going at ~40s with empty output, detach or `stop_job` — piping to `head` does not help.\n\n## PR checks (only if a goal is given)\n\nDo **not** watch after every push. If the user gave a goal (Greptile 5\u002F5, CI green, until checks pass), wait the same way:\n\n```bash\nnode scripts\u002Fdetached-job.cjs start pr-checks -- gh pr checks --watch\n```\n\nOn miss, fix, commit, push, and watch again until the goal is met or you are blocked. Do not ask the human to poll. Do not merge unless asked.\n\n## Mac desktop release\n\nDefault: push\u002Fretarget the `v*` tag and watch Actions — do not pack locally. Full steps: [`.claude\u002Fskills\u002Frelease\u002FSKILL.md`](..\u002Frelease\u002FSKILL.md).\n\n```bash\nnode scripts\u002Fdetached-job.cjs start gha-release -- node scripts\u002Fwatch-gha-run.cjs \u003Crun-id>\nnode scripts\u002Fdetached-job.cjs wait gha-release\n```\n\nDo **not** `gh run watch` (full TTY reprint every 3s; stalls the log pane).\n","# Long-running jobs\n\nSideboard ships this method to **every** worktree agent (`\u002Flong-running` is a product skill). This repo copy adds Mac-release notes below. A workspace skill with the same name wins over the bundled copy.\n\nA Sideboard or Cursor **worktree turn** SIGTERMs the agent shell (and its process group) when the user sends another message or the turn is interrupted. `block_until_ms: 0` is not enough — the child stays in that group.\n\nDo **not** ask the human to check back. Detach, then **wait** in 45s slices (same idea as MCP `wait_for_turn`) until `stillRunning` is false. Do not end the turn after starting a job — a native CLI Task or bash notify does **not** resume a Sideboard chat (Claude, Cursor, Codex, OpenCode).\n\nA process that is supposed to stay up is not this skill. Do not `start` a static file server, preview server, or anything else that never exits (`python -m http.server`, a dev server). `stillRunning` stays true, and the orchestrator will not let the turn end. To show the local marketing page, open `site\u002Findex.html` (side column or a browser). If a stay-up server is already a detached job, `stop` that id and continue. Do not loop `wait` on it.\n\n## Tool\n\nPrefer the helper path from the Sideboard playbook (absolute `node \"…\" start`). In this repo, `scripts\u002Fdetached-job.cjs` is the same file.\n\n```bash\n# Start (exits in ~1s; job survives this turn)\nnode scripts\u002Fdetached-job.cjs start \u003Cid> -- \u003Ccommand> [args...]\n\n# Wait — returns within ~45s even if the job is still going\nnode scripts\u002Fdetached-job.cjs wait \u003Cid>\n\n# Block until the process exits (humans \u002F a turn that will not be interrupted)\nnode scripts\u002Fdetached-job.cjs wait \u003Cid> --until-done\n\nnode scripts\u002Fdetached-job.cjs stop \u003Cid> [--reason TEXT]\nnode scripts\u002Fdetached-job.cjs status \u003Cid>\nnode scripts\u002Fdetached-job.cjs ui \u003Cid> [--title TEXT] [--out FILE]\n```\n\n`\u003Cid>` is kebab-case (`mac-release`, `core-test`, `fly-deploy`). State is `.context\u002F.sideboard\u002Fdetached-jobs\u002F\u003Cid>\u002F` (local scratch).\n\nWait JSON:\n\n- `stillRunning: true` → exit 2 → **call wait again**. Progress is in `progress` \u002F `phase`. Do not start a second job. Do not ping the user. If it is hanging, producing no useful output, or doing the wrong thing, call MCP `stop_job` (or `detached-job.cjs stop`) instead of looping forever.\n- `ok: true` → exit 0 → continue the rest of the task (README, commit, deploy next step).\n- `failed: true` → exit 1 → read `progress`, fix, start **once**.\n- `stopped: true` → you ended it. Present the last delta (`status=failed`) and decide the next step.\n\n## Sideboard UI (append-only log)\n\nThe side column **is** the live view. `wait_for_job` \u002F `stop_job` (and shell `detached-job` wait JSON) open a `type=log` pane automatically (`artifact_id` = job id, content = `delta`). You may still call `present_artifact` **`type=log`** for a custom title. Each update **appends** — send only `delta`, not the full log and not HTML.\n\n```\npresent_artifact\n  title: \u003Cshort title>\n  type: log\n  artifact_id: \u003Cid>\n  content: \u003Cwait.delta>          # new lines only; empty is ok\n  status: running | ok | failed  # from wait.status\n  phase: \u003Cwait.phase>\n```\n\nAfter **start**, the first `wait` opens the pane (`status=running`). Do **not** dump the log in chat. Do **not** also fence HTML.\n\n`detached-job.cjs ui` still writes a snapshot HTML file for debugging. Prefer `type=log`.\n\nIf you started via `release-mac-detached.js` (legacy pid\u002Flog):\n\n```bash\nnode scripts\u002Fdetached-job.cjs ui --pid-file apps\u002Fdesktop\u002Frelease\u002Frelease.pid --log-file apps\u002Fdesktop\u002Frelease\u002Frelease.log --ok-pattern RELEASE_BUILD_OK --title \"Mac pack\"\n```\n\nAd-hoc pid\u002Flog (legacy or another tool’s files):\n\n```bash\nnode scripts\u002Fdetached-job.cjs wait --pid-file FILE --log-file FILE [--ok-pattern TEXT]\n```\n\n## Agent loop\n\n1. `start` once. If JSON says `already-running`, do not start again.\n2. Loop `wait` (use `--timeout-ms 15000` for a livelier pane). The human should see **working** in the side column from the wait JSON — do not also dump the log in chat.\n3. After each slice, the same log pane appends `delta`. Optional: `present_artifact` the same id for a custom title.\n4. On `ok`, present once more (`status=ok`, last `delta`) and finish the task. On `failed`, fix from the log.\n5. **Stop** when the job is hanging, buffering forever, watching the wrong thing, or you already have the answer. MCP `stop_job` (same id, optional `reason`) or `node scripts\u002Fdetached-job.cjs stop \u003Cid>`. Do **not** stop a pack\u002Ftest\u002Fdeploy that is clearly making progress.\n\nNever tell the user “say status when it’s done.” You wait — unless you decided to stop.\n\n## Connector CLIs (Vercel, Supabase, Sentry, PostHog)\n\nHuge `--json` \u002F `--expand` dumps crash any worktree agent mid-turn (Claude \u002F Cursor \u002F Codex \u002F OpenCode). On Cursor the chat shows the packaged bundle (`file:\u002F\u002F…\u002Fcursor-runtime\u002F…\u002F@cursor\u002Fsdk\u002Fdist\u002Fesm\u002Findex.js` then minified `importas e from\"@bufbuild\u002Fprotobuf\"`). Write CLI\u002FHTTP output to `.context\u002Fcli\u002F` (not `.context\u002Fattachments\u002F`), then read a slice. Tight window first; one query at a time; `jq` only the fields you need. After the error line, stop fetching. If a raw shell is still going at ~40s with empty output, detach or `stop_job` — piping to `head` does not help.\n\n## PR checks (only if a goal is given)\n\nDo **not** watch after every push. If the user gave a goal (Greptile 5\u002F5, CI green, until checks pass), wait the same way:\n\n```bash\nnode scripts\u002Fdetached-job.cjs start pr-checks -- gh pr checks --watch\n```\n\nOn miss, fix, commit, push, and watch again until the goal is met or you are blocked. Do not ask the human to poll. Do not merge unless asked.\n\n## Mac desktop release\n\nDefault: push\u002Fretarget the `v*` tag and watch Actions — do not pack locally. Full steps: [`.claude\u002Fskills\u002Frelease\u002FSKILL.md`](..\u002Frelease\u002FSKILL.md).\n\n```bash\nnode scripts\u002Fdetached-job.cjs start gha-release -- node scripts\u002Fwatch-gha-run.cjs \u003Crun-id>\nnode scripts\u002Fdetached-job.cjs wait gha-release\n```\n\nDo **not** `gh run watch` (full TTY reprint every 3s; stalls the log pane).\n",{"name":5,"description":7},"main","02fe9a8780abdb3554a042bd4e4d84b774167ee7",".claude\u002Fskills\u002Flong-running",0,"https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard","https:\u002F\u002Fgithub.com\u002Fmattlevine","2026-11-15 03:17:37",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add mattlevine\u002Fsideboard --skill long-running",{"tool":37,"command":38},"bunx","bunx skills add mattlevine\u002Fsideboard --skill long-running",{"tool":40,"command":41},"pnpm","pnpm dlx skills add mattlevine\u002Fsideboard --skill long-running","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Flong-running\u002FSKILL.md",true,[],[46,53,59,64,71,78],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":19},"01M7E0618TWSY135JSVJYDSDB7","skill-improver","When a miss will recur — you made it twice, or a person told you the rule — write the lesson into the skill before the turn ends. One sentence, in the same change. Run in the turn, not on a schedule.","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskill-improver\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"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":56,"categories":57,"keywords":58,"publishedAt":19},"01M7A0C7G870SK6SZTGMATAF26","release","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.cursor\u002Fskills\u002Frelease\u002FSKILL.md",[],[],{"skillNo":60,"name":5,"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":61,"categories":62,"keywords":63,"publishedAt":19},"01M7A0C5ABGQBF6Q94FSXH6Y2M","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.cursor\u002Fskills\u002Flong-running\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7A0C2YWWYF5PME33KYS13SP","worktree-install","Reinstall dependencies in a Sideboard worktree after merge-from-main or a lockfile change. Use when `pnpm install` hangs, several worktrees install at once, or a native build is outside the permitted directory.","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fworktree-install\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7A0BWTS631GV69KGJW80RDQ","review","Merge-readiness review for this repository. Use when reviewing a PR, worktree diff, or Sideboard Review (Approve \u002F Approve with nits \u002F Request changes).","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Freview\u002FSKILL.md",[],[],{"skillNo":79,"name":55,"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},"01M7A0BDG6EZP7GTVVT0Z8FFS0","Cut a Sideboard version. Default: bump, README\u002Fchangelog, commit, merge, then push\u002Fretarget the `v*` tag so `.github\u002Fworkflows\u002Frelease.yml` publishes npm (OIDC) and the signed Mac desktop. No workflow_dispatch. Do not pack Electron in a worktree turn. Use for electron release, desktop release, pnpm release, or “I’ll publish npms.”","https:\u002F\u002Fgithub.com\u002Fmattlevine\u002Fsideboard\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Frelease\u002FSKILL.md",[],[]]