[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6AD7862D8WH4576CB754ZQ5":3,"skill-same:01M6AD7862D8WH4576CB754ZQ5":78},{"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":23,"publishedAt":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":41,"commitSha":42,"skillDirectory":43,"forksCount":44,"repositoryHtmlUrl":45,"repositoryHomepageUrl":6,"ownerHtmlUrl":46,"ownerBlogUrl":47,"ownerDescription":48,"updatedAt":49,"install":50,"resources":63},"01M6AD7862D8WH4576CB754ZQ5","steer-once",null,"Turn a correction you keep repeating into a durable fix: mine your corrections, classify them, choose the cheapest layer that would have prevented them (instruction, skill, lint\u002FCI check, test, automation, or product bug report), open a PR with an evidence block, then re-measure after K days and keep, escalate or revert. Opens PRs, so it runs only on `\u002Flever steer` or an explicit ask (\"mine my corrections\", \"why do I keep telling you X\", \"didn't I give you a skill for this\"), or when prompt-mirror reports the same correction class in 3+ sessions and you say go.","github\u002Flever","github","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9919?v=4","01M6AC83DS133TPXFSP1DHNCC6","GitHub","01M6AD6AQD82NPA8BVHR3MZKS2",289,"MIT","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Fsteer-once\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35],"steer once","go","automation","test","sessions","pr","bug report","lint","report","product","ci","lever","2026-11-10 10:12:30",false,"---\nname: steer-once\nuser-invocable: false\ndescription: >-\n  Turn a correction you keep repeating into a durable fix: mine your\n  corrections, classify them, choose the cheapest layer that would have\n  prevented them (instruction, skill, lint\u002FCI check, test, automation, or\n  product bug report), open a PR with an evidence block, then re-measure after\n  K days and keep, escalate or revert. Opens PRs, so it runs only on\n  `\u002Flever steer` or an explicit ask (\"mine my corrections\", \"why do I keep\n  telling you X\", \"didn't I give you a skill for this\"), or when prompt-mirror\n  reports the same correction class in 3+ sessions and you say go.\n---\n\n# steer-once\n\nCorrect an agent once; the repo remembers. An apology from the agent changes nothing next session. A merged rule does. This skill takes corrections you have typed more than once, finds the layer that would have prevented them, lands the smallest durable change there with evidence, and checks a week later whether the correction actually stopped. If it didn't, it climbs one layer; if the rule made things worse, it proposes a revert.\n\n## When to use\n\n- **Fast path (live):** the user just repeated a correction mid-session (\"again\", \"I told you\", \"why did you stop\", \"didn't I give you a skill\"). Handle that one correction now.\n- **Batch:** \"mine my corrections\", a weekly retro, or a `prompt-mirror` report listing a class with ≥N hits across ≥3 sessions.\n- **Team variant:** recurring reviewer feedback on agent PRs in a repo (see Sources).\n\nNot for one-off preferences, a single bad turn, or anything already handled by an open steer-once PR.\n\n## Sources\n\n1. **Your typed messages** via [`prompt-mirror`](..\u002Fprompt-mirror\u002FSKILL.md)'s extractor (local, redacted, source-filtered):\n   `python3 \u003Cskills-dir>\u002Fprompt-mirror\u002Fscripts\u002Fmine_sessions.py --since 21d [--repo owner\u002Fname]`.\n   Use `correction_classes`, `repeated_correction_phrases`, and each message's `classes`\u002F`flags`. Read `messages.jsonl` for the instances.\n2. **The agent message before each correction** (the cause) and **after** (the response). Count concession openers (\"You're right\", \"Agreed\", \"Good catch\") on the same class; a concession followed by a repeat is the strongest signal that a rule is missing or not loaded.\n3. **Team variant:** `changes_requested` reviews and review comments on agent-authored PRs (for example `gh pr list --author app\u002Fcopilot-swe-agent --state all`, then `gh api repos\u002F{owner}\u002F{repo}\u002Fpulls\u002F{n}\u002Freviews`). Same pipeline, no local history needed.\n\n## Steps\n\n1. **Read the dials** (table below) and [`docs\u002Fautonomy.md`](..\u002F..\u002Fdocs\u002Fautonomy.md). State the scope and autonomy level in one line.\n2. **Collect** corrections for the window. Exclude agent, autopilot, system and template traffic (the extractor does this). Keep delivery mode: a mid-turn steer is a stronger correction than an idle follow-up.\n3. **Classify** each correction. Start from the extractor's classes, then confirm by reading:\n\n   | Class | The agent… | Typical correction |\n   |---|---|---|\n   | gate | invented approvals, locks, preflights, \"exact GO\" phrases | \"you don't need authorization\" |\n   | over_ask | asked for what it already had | \"don't ask, just do it\" |\n   | premature_stop | treated a milestone as done | \"why did you stop?\" |\n   | polling | waited or slept instead of doing parallel work | \"why are you waiting?\" |\n   | status_gap | made the human poll for state | \"status? ETA?\" |\n   | verification_theatre | let checking replace shipping | \"what did you spend 3 hours on?\" |\n   | stale_claim | reasoned from a proxy (stale clone, cache) | \"that's not true, check the live source\" |\n   | shallow_fix | suppressed the symptom | \"actually fix it instead of disabling lint\" |\n   | letter_of_law | shrank an ambitious ask | \"I asked for the orchestrator pattern\" |\n   | scope_creep | did unrequested side quests | \"why did you open a PR there?\" |\n   | process_default | lacked a team default procedure | \"breaking changes are always two-phase\" |\n   | runtime_gap | hit something no prompt can fix | \"rekick the stalled sessions\", quota, leaked processes |\n\n4. **Find recurrences:** same class and same underlying expectation, ≥N hits across ≥2 sessions in the window (default N=3). Merge near-duplicates; one rule per expectation.\n5. **Check for an existing rule.** Search `.github\u002Fcopilot-instructions.md`, `.github\u002Finstructions\u002F*.instructions.md` (`applyTo` scope), `AGENTS.md`, `.github\u002Fskills\u002F*\u002FSKILL.md`, `.github\u002Flever.yml`, and personal `~\u002F.copilot\u002Fcopilot-instructions.md`. If a rule exists and the class still recurs, the finding is **\"rule exists but is weak, misplaced or not loaded\"**. Fix placement or wording, or escalate a layer, before adding more prose.\n6. **Pick the cheapest durable layer** that would have prevented these specific instances:\n\n   | The expectation is… | Layer | Example |\n   |---|---|---|\n   | A personal posture (tone, verbosity, how you like updates) | Personal instructions | \"Report as `x → y (Δz%)`; no preamble\" |\n   | A repo-wide default posture | `.github\u002Fcopilot-instructions.md` (one line, with the why) | \"Pre-launch service: ship and monitor; don't add approval gates\" |\n   | Authority or autonomy | A dial in `.github\u002Flever.yml`, not prose | `autonomy: L3` with the grant recorded |\n   | A task-specific procedure | Scoped instruction or skill rule | expand → migrate → contract for breaking changes |\n   | Mechanically detectable | Lint or CI check | no new `eslint-disable` without a linked issue |\n   | Behavioural | A test that fails today | the service recovers from a stuck lease by itself |\n   | Timing or liveness | Hook or automation (see [`automate-it`](..\u002Fautomate-it\u002FSKILL.md)) | resume idle child sessions on wake |\n   | Outside the agent's control | Product or infra bug report | sessions don't auto-resume after sleep |\n\n   Corrections that cluster on one dial (merge authority, stop policy, status cadence, validation depth) become a dial change, not a new rule.\n7. **Write the rule.** Positive, concrete, with the why and the scope, at most 3 lines. Name the dial it sets. No ALL CAPS, no \"NEVER\" walls, and never add a gate to prevent gates. Example: \"Autopilot runs never idle: when the current item lands, take the next ready item. Why: 'why did you stop' recurred in 4 sessions.\"\n8. **Open the PR** (or file the issue, for bugs) with the evidence block below. For personal instructions there is no PR: show the diff and apply only when the user says so, or under a recorded grant. Show the user the exact evidence text before it is published to any shared repo.\n9. **Land per the autonomy dial.** Default L2: a ready PR the human lands. Instruction-only PRs are a common L3 grant class (\"docs and instructions PRs may self-merge when green\"). Record which level applied.\n10. **Record** in the ledger: class, rule, layer, PR, baseline rate, re-measure date.\n11. **Re-measure after K days** (and at least M typed messages in scope). Re-run the extractor for the same scope. Compare the class rate per 100 typed messages with the baseline, and with the week-to-week spread of that class over the previous 4 weeks.\n    - Fell beyond noise → **keep**; comment the result on the PR.\n    - Didn't fall → **escalate one layer** (instruction → skill rule → check\u002Ftest → automation → bug report) and link the new PR to the old one.\n    - A neighbour class rose (for example over_ask fell but premature_stop rose) → **propose a revert** or rewording, with numbers.\n\n## Evidence block (PR body)\n\n```markdown\n### steer-once evidence\n- **Class:** premature_stop: agent treated a milestone as done while ready work remained\n- **Window \u002F scope:** 2026-10-01 → 2026-10-21, this repo, typed messages only\n- **Count:** 9 corrections in 4 sessions = 1.6 per 100 typed messages (baseline)\n- **Agent concessions on this class:** 5 (\"You're right…\"), followed by repeats\n- **Existing rule:** none | `path:line` exists but recurred (placement\u002Fwording issue)\n- **Quotes** (redacted, dated, ≤20 words each, 3–5):\n  - 2026-10-04 \"why did you stop? there is more in the queue\"\n  - 2026-10-09 \"if you're idle, take the next ready item\"\n- **Layer:** repo instruction. Cheaper layer insufficient because: … Costlier layer unnecessary because: …\n- **Keep bar:** rate ≤ 50% of baseline and outside 4-week noise, re-measured on 2026-10-28 (K=7) or after 150 typed msgs\n- **If not met:** escalate to \u003Cnext layer>\n```\n\n## Dials\n\n| Dial | Default | Options |\n|---|---|---|\n| Detection threshold | N=3 hits across ≥2 sessions in 21 days | any N \u002F window |\n| Mode | open PR (L2, human lands) | propose in chat; draft PR (L1); self-merge per [`docs\u002Fautonomy.md`](..\u002F..\u002Fdocs\u002Fautonomy.md) grant |\n| Destination | repo for shared expectations, personal for posture | repo only; personal only |\n| Evidence | redacted quotes + counts, shown to user before publishing | counts only; paraphrase |\n| Re-measure | K=7 days and ≥150 typed messages in scope | any K \u002F M |\n| Keep bar | ≤50% of baseline and outside noise | any threshold |\n| Max open steer-once PRs | 3 | any |\n| Severity weighting | off | weight mid-turn steers and heated messages higher (never quote heated ones) |\n\n## Outputs\n\n- One small PR or issue per recurring expectation, labelled `steer-once`, with the evidence block. The label is the team ledger.\n- A personal ledger entry (default `~\u002F.copilot\u002Flever\u002Fsteer-once.ledger.md`, format per [`templates\u002Fledger.md`](..\u002F..\u002Ftemplates\u002Fledger.md)): keep, escalate or revert, with before\u002Fafter rates.\n- A re-measure comment on the original PR.\n\n## This skill's own feedback loop\n\nPer rule: class rate per 100 typed messages, before and after. Across rules: share kept versus escalated versus reverted; concession openers per 100 agent replies (should fall); time from first occurrence to merged fix (should shrink). If most rules escalate, the skill is choosing layers too cheaply. If any rule raises a neighbour class, the skill is writing over-broad rules.\n\n## Failure modes it prevents\n\n- Accepting \"You're right\" as a fix.\n- Prose rules for things a check, test, dial or bug report should own.\n- Rule sprawl: duplicate, contradictory or ALL-CAPS instructions nobody re-measures.\n- One person's posture hard-coded into a shared repo; posture goes to personal instructions or a dial.\n- Private text published as \"evidence\" without the user seeing it first.\n\n## Why this exists\n\n- Across one heavy user's sessions, agents opened 75 replies with \"You're right\" or an equivalent, and the same corrections recurred for days. They stopped recurring only once they became a merged instruction or skill change.\n- The fastest path, from a frustrated message to a merged operating-model PR, took 16 minutes. One correction that stayed verbal took about 18 days to become a default.\n- After the user had an agent mine his last ~200 typed messages into a skill rewrite: \"why did you stop\" nudges 27 → 0, gate language in agent messages 31% → 3%, status asks ~24% → 7%, while his message volume rose from about 50 to 87 per day. \"The delivery skill is ~200 of my own corrections, compressed.\"\n- Some corrections are bug reports. \"Restart the stalled sessions\" recurred 8 times after the instruction existed, because interrupted sessions were a runtime gap no prompt could fix.\n","# steer-once\n\nCorrect an agent once; the repo remembers. An apology from the agent changes nothing next session. A merged rule does. This skill takes corrections you have typed more than once, finds the layer that would have prevented them, lands the smallest durable change there with evidence, and checks a week later whether the correction actually stopped. If it didn't, it climbs one layer; if the rule made things worse, it proposes a revert.\n\n## When to use\n\n- **Fast path (live):** the user just repeated a correction mid-session (\"again\", \"I told you\", \"why did you stop\", \"didn't I give you a skill\"). Handle that one correction now.\n- **Batch:** \"mine my corrections\", a weekly retro, or a `prompt-mirror` report listing a class with ≥N hits across ≥3 sessions.\n- **Team variant:** recurring reviewer feedback on agent PRs in a repo (see Sources).\n\nNot for one-off preferences, a single bad turn, or anything already handled by an open steer-once PR.\n\n## Sources\n\n1. **Your typed messages** via [`prompt-mirror`](..\u002Fprompt-mirror\u002FSKILL.md)'s extractor (local, redacted, source-filtered):\n   `python3 \u003Cskills-dir>\u002Fprompt-mirror\u002Fscripts\u002Fmine_sessions.py --since 21d [--repo owner\u002Fname]`.\n   Use `correction_classes`, `repeated_correction_phrases`, and each message's `classes`\u002F`flags`. Read `messages.jsonl` for the instances.\n2. **The agent message before each correction** (the cause) and **after** (the response). Count concession openers (\"You're right\", \"Agreed\", \"Good catch\") on the same class; a concession followed by a repeat is the strongest signal that a rule is missing or not loaded.\n3. **Team variant:** `changes_requested` reviews and review comments on agent-authored PRs (for example `gh pr list --author app\u002Fcopilot-swe-agent --state all`, then `gh api repos\u002F{owner}\u002F{repo}\u002Fpulls\u002F{n}\u002Freviews`). Same pipeline, no local history needed.\n\n## Steps\n\n1. **Read the dials** (table below) and [`docs\u002Fautonomy.md`](..\u002F..\u002Fdocs\u002Fautonomy.md). State the scope and autonomy level in one line.\n2. **Collect** corrections for the window. Exclude agent, autopilot, system and template traffic (the extractor does this). Keep delivery mode: a mid-turn steer is a stronger correction than an idle follow-up.\n3. **Classify** each correction. Start from the extractor's classes, then confirm by reading:\n\n   | Class | The agent… | Typical correction |\n   |---|---|---|\n   | gate | invented approvals, locks, preflights, \"exact GO\" phrases | \"you don't need authorization\" |\n   | over_ask | asked for what it already had | \"don't ask, just do it\" |\n   | premature_stop | treated a milestone as done | \"why did you stop?\" |\n   | polling | waited or slept instead of doing parallel work | \"why are you waiting?\" |\n   | status_gap | made the human poll for state | \"status? ETA?\" |\n   | verification_theatre | let checking replace shipping | \"what did you spend 3 hours on?\" |\n   | stale_claim | reasoned from a proxy (stale clone, cache) | \"that's not true, check the live source\" |\n   | shallow_fix | suppressed the symptom | \"actually fix it instead of disabling lint\" |\n   | letter_of_law | shrank an ambitious ask | \"I asked for the orchestrator pattern\" |\n   | scope_creep | did unrequested side quests | \"why did you open a PR there?\" |\n   | process_default | lacked a team default procedure | \"breaking changes are always two-phase\" |\n   | runtime_gap | hit something no prompt can fix | \"rekick the stalled sessions\", quota, leaked processes |\n\n4. **Find recurrences:** same class and same underlying expectation, ≥N hits across ≥2 sessions in the window (default N=3). Merge near-duplicates; one rule per expectation.\n5. **Check for an existing rule.** Search `.github\u002Fcopilot-instructions.md`, `.github\u002Finstructions\u002F*.instructions.md` (`applyTo` scope), `AGENTS.md`, `.github\u002Fskills\u002F*\u002FSKILL.md`, `.github\u002Flever.yml`, and personal `~\u002F.copilot\u002Fcopilot-instructions.md`. If a rule exists and the class still recurs, the finding is **\"rule exists but is weak, misplaced or not loaded\"**. Fix placement or wording, or escalate a layer, before adding more prose.\n6. **Pick the cheapest durable layer** that would have prevented these specific instances:\n\n   | The expectation is… | Layer | Example |\n   |---|---|---|\n   | A personal posture (tone, verbosity, how you like updates) | Personal instructions | \"Report as `x → y (Δz%)`; no preamble\" |\n   | A repo-wide default posture | `.github\u002Fcopilot-instructions.md` (one line, with the why) | \"Pre-launch service: ship and monitor; don't add approval gates\" |\n   | Authority or autonomy | A dial in `.github\u002Flever.yml`, not prose | `autonomy: L3` with the grant recorded |\n   | A task-specific procedure | Scoped instruction or skill rule | expand → migrate → contract for breaking changes |\n   | Mechanically detectable | Lint or CI check | no new `eslint-disable` without a linked issue |\n   | Behavioural | A test that fails today | the service recovers from a stuck lease by itself |\n   | Timing or liveness | Hook or automation (see [`automate-it`](..\u002Fautomate-it\u002FSKILL.md)) | resume idle child sessions on wake |\n   | Outside the agent's control | Product or infra bug report | sessions don't auto-resume after sleep |\n\n   Corrections that cluster on one dial (merge authority, stop policy, status cadence, validation depth) become a dial change, not a new rule.\n7. **Write the rule.** Positive, concrete, with the why and the scope, at most 3 lines. Name the dial it sets. No ALL CAPS, no \"NEVER\" walls, and never add a gate to prevent gates. Example: \"Autopilot runs never idle: when the current item lands, take the next ready item. Why: 'why did you stop' recurred in 4 sessions.\"\n8. **Open the PR** (or file the issue, for bugs) with the evidence block below. For personal instructions there is no PR: show the diff and apply only when the user says so, or under a recorded grant. Show the user the exact evidence text before it is published to any shared repo.\n9. **Land per the autonomy dial.** Default L2: a ready PR the human lands. Instruction-only PRs are a common L3 grant class (\"docs and instructions PRs may self-merge when green\"). Record which level applied.\n10. **Record** in the ledger: class, rule, layer, PR, baseline rate, re-measure date.\n11. **Re-measure after K days** (and at least M typed messages in scope). Re-run the extractor for the same scope. Compare the class rate per 100 typed messages with the baseline, and with the week-to-week spread of that class over the previous 4 weeks.\n    - Fell beyond noise → **keep**; comment the result on the PR.\n    - Didn't fall → **escalate one layer** (instruction → skill rule → check\u002Ftest → automation → bug report) and link the new PR to the old one.\n    - A neighbour class rose (for example over_ask fell but premature_stop rose) → **propose a revert** or rewording, with numbers.\n\n## Evidence block (PR body)\n\n```markdown\n### steer-once evidence\n- **Class:** premature_stop: agent treated a milestone as done while ready work remained\n- **Window \u002F scope:** 2026-10-01 → 2026-10-21, this repo, typed messages only\n- **Count:** 9 corrections in 4 sessions = 1.6 per 100 typed messages (baseline)\n- **Agent concessions on this class:** 5 (\"You're right…\"), followed by repeats\n- **Existing rule:** none | `path:line` exists but recurred (placement\u002Fwording issue)\n- **Quotes** (redacted, dated, ≤20 words each, 3–5):\n  - 2026-10-04 \"why did you stop? there is more in the queue\"\n  - 2026-10-09 \"if you're idle, take the next ready item\"\n- **Layer:** repo instruction. Cheaper layer insufficient because: … Costlier layer unnecessary because: …\n- **Keep bar:** rate ≤ 50% of baseline and outside 4-week noise, re-measured on 2026-10-28 (K=7) or after 150 typed msgs\n- **If not met:** escalate to \u003Cnext layer>\n```\n\n## Dials\n\n| Dial | Default | Options |\n|---|---|---|\n| Detection threshold | N=3 hits across ≥2 sessions in 21 days | any N \u002F window |\n| Mode | open PR (L2, human lands) | propose in chat; draft PR (L1); self-merge per [`docs\u002Fautonomy.md`](..\u002F..\u002Fdocs\u002Fautonomy.md) grant |\n| Destination | repo for shared expectations, personal for posture | repo only; personal only |\n| Evidence | redacted quotes + counts, shown to user before publishing | counts only; paraphrase |\n| Re-measure | K=7 days and ≥150 typed messages in scope | any K \u002F M |\n| Keep bar | ≤50% of baseline and outside noise | any threshold |\n| Max open steer-once PRs | 3 | any |\n| Severity weighting | off | weight mid-turn steers and heated messages higher (never quote heated ones) |\n\n## Outputs\n\n- One small PR or issue per recurring expectation, labelled `steer-once`, with the evidence block. The label is the team ledger.\n- A personal ledger entry (default `~\u002F.copilot\u002Flever\u002Fsteer-once.ledger.md`, format per [`templates\u002Fledger.md`](..\u002F..\u002Ftemplates\u002Fledger.md)): keep, escalate or revert, with before\u002Fafter rates.\n- A re-measure comment on the original PR.\n\n## This skill's own feedback loop\n\nPer rule: class rate per 100 typed messages, before and after. Across rules: share kept versus escalated versus reverted; concession openers per 100 agent replies (should fall); time from first occurrence to merged fix (should shrink). If most rules escalate, the skill is choosing layers too cheaply. If any rule raises a neighbour class, the skill is writing over-broad rules.\n\n## Failure modes it prevents\n\n- Accepting \"You're right\" as a fix.\n- Prose rules for things a check, test, dial or bug report should own.\n- Rule sprawl: duplicate, contradictory or ALL-CAPS instructions nobody re-measures.\n- One person's posture hard-coded into a shared repo; posture goes to personal instructions or a dial.\n- Private text published as \"evidence\" without the user seeing it first.\n\n## Why this exists\n\n- Across one heavy user's sessions, agents opened 75 replies with \"You're right\" or an equivalent, and the same corrections recurred for days. They stopped recurring only once they became a merged instruction or skill change.\n- The fastest path, from a frustrated message to a merged operating-model PR, took 16 minutes. One correction that stayed verbal took about 18 days to become a default.\n- After the user had an agent mine his last ~200 typed messages into a skill rewrite: \"why did you stop\" nudges 27 → 0, gate language in agent messages 31% → 3%, status asks ~24% → 7%, while his message volume rose from about 50 to 87 per day. \"The delivery skill is ~200 of my own corrections, compressed.\"\n- Some corrections are bug reports. \"Restart the stalled sessions\" recurred 8 times after the instruction existed, because interrupted sessions were a runtime gap no prompt could fix.\n",{"name":5,"description":7,"user-invocable":37},"main","f859d0aa2be96981157ee0480655c84877e3cd98","skills\u002Fsteer-once",14,"https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever","https:\u002F\u002Fgithub.com\u002Fgithub","https:\u002F\u002Fgithub.com\u002Fabout","How people build software.","2026-11-06 22:44:16",{"source":8,"skillName":5,"commands":51,"prompt":61,"zipAvailable":62},[52,55,58],{"tool":53,"command":54},"npx","npx skills add github\u002Flever --skill steer-once",{"tool":56,"command":57},"bunx","bunx skills add github\u002Flever --skill steer-once",{"tool":59,"command":60},"pnpm","pnpm dlx skills add github\u002Flever --skill steer-once","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Fsteer-once\u002FSKILL.md",true,[64,70,74],{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Fimport_allowlist.py","SCRIPT","脚本",9648,"https:\u002F\u002Fgithub.com\u002Fshengdabai\u002FTony-Claude-Code-Skills\u002Fblob\u002Fmain\u002Fskills\u002Fdesign-token-ledger\u002Fscripts\u002Fimport_allowlist.py",{"path":71,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":72,"sourceHtmlUrl":73},"scripts\u002Ftoken_ledger.py",23398,"https:\u002F\u002Fgithub.com\u002Fshengdabai\u002FTony-Claude-Code-Skills\u002Fblob\u002Fmain\u002Fskills\u002Fdesign-token-ledger\u002Fscripts\u002Ftoken_ledger.py",{"path":75,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":76,"sourceHtmlUrl":77},"tests\u002Ftest_gates.py",9614,"https:\u002F\u002Fgithub.com\u002Fshengdabai\u002FTony-Claude-Code-Skills\u002Fblob\u002Fmain\u002Fskills\u002Fdesign-token-ledger\u002Ftests\u002Ftest_gates.py",[79,91,106,117,134,147],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":36},"01M6AD767J9CB04Y0FY7JFSKJJ","run-health-review","Stall watchdog and automation health review. Pulls recent runs, merges, no-op rate, failures, rate-limit hits and stuck PRs; classifies each stall; proposes exactly one change or \"no change\"; and takes over orphaned work. Use when unattended delivery stalled (\"why did it stall overnight\", \"no merges in N hours\", \"it was fast the first few days, now slow\"), when an automation fails or its output drops, when rate limits or quotas are hit, or on a schedule while agents run unattended.","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Frun-health-review\u002FSKILL.md",[],[86,87,88,89,90,26],"run health review","review","run","fast","change",{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":100,"publishedAt":36},"01M6AD746F7K4FQG99EE8CZ6DT","refactor-pipeline","Runs a huge, mechanical refactor (crate or package splits, module extraction, API migrations, dead-code removal) as three independent lanes, producer, packager and shepherd, that coordinate only through git and a PR stack. It pushes branches and opens PRs, so it runs only on `\u002Flever refactor` or an explicit ask (\"split this crate\", \"break up this monolith\", \"this module is huge, cut it down\", \"migrate every caller\").","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Frefactor-pipeline\u002FSKILL.md",[97],{"categoryNo":98,"categoryName":99,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000007","代码审查",[101,29,102,103,104,105,35],"refactor pipeline","code","git","package","extraction",{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":113,"publishedAt":36},"01M6AD72CGZ51VEDN1PZZPB7VB","prompt-mirror","Mine your own local Copilot session history and return a scorecard of how you direct agents, plus three suggestions illustrated with your own before\u002Fafter messages. It reads your private history, so it runs only on `\u002Flever mirror` or an explicit ask (\"how am I prompting?\", \"coach me\", \"mine my sessions\", \"what do I keep repeating?\", a weekly self-retro).","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Fprompt-mirror\u002FSKILL.md",[112],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[114,28,115,116,35],"prompt mirror","copilot","messages",{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":128,"publishedAt":36},"01M6AD70V6MPP4EE686C6TR4MG","progress-contract","Pushes progress on long-running or delegated work so the human never has to poll. It publishes milestones with ETAs and the evidence each needs, sends reports only when something meaningful changed, labels every claim with an evidence level, logs ETA accuracy, watches time since last merge, and audits \"done\" against evidence before declaring it. Use for any delegated, multi-hour or overnight run, when a brief sets a report contract, or when the user asks \"status?\", \"ETA?\", \"where are we?\", \"what happened overnight?\", \"hour by hour\" or \"are you done?\".","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Fprogress-contract\u002FSKILL.md",[123],{"categoryNo":124,"categoryName":125,"taxonomyType":126,"taxonomyTypeDesc":127},"01JBSEEDCAT000000000000037","律师与法务","PROFESSION","适用职业",[129,130,131,88,32,132,133],"progress contract","logs","brief","merge","status",{"skillNo":135,"name":136,"nameZh":6,"description":137,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":138,"categories":139,"keywords":140,"publishedAt":36},"01M6AD6YMC170PX6A5VWDE4T2B","pr-shepherd","Drives an open pull request to landed through an event-driven loop. On each wake it does one batched read, fixes root causes of red required checks, dispositions every review thread, and lands at the level set by the autonomy dial. It pushes and can merge, so it runs only on `\u002Flever shepherd`, a hand-off from a lever coordinator, a PR labeled for shepherding, or an explicit ask (\"monitor CI until green\", \"push fixes until green\", \"resolve the review threads\", \"merge when green\", \"babysit this PR\", \"shepherd","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Fpr-shepherd\u002FSKILL.md",[],[141,29,142,87,132,143,144,145,34,146,35],"pr shepherd","pull request","monitor","pull","push","autonomy dial",{"skillNo":148,"name":35,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":152,"publishedAt":36},"01M6AD6WNP2AN43C3HG4MBJH8B","The one lever command. `\u002Flever \u003Cverb> [args]` runs a lever skill directly (loop, arms, review, brief, deliver, refactor, backlog, shepherd, land, status, automate, health, mirror, steer, autonomy, spend, setup, help). `\u002Flever \u003Cplain words>` picks the verb for you, says which in one line, then runs it. Also reads and writes the user's personal dials in `~\u002F.copilot\u002Flever.yml`. Use when the user types \u002Flever, asks \"which lever skill should I use?\", or wants to see or change their autonomy or spend dial.","https:\u002F\u002Fgithub.com\u002Fgithub\u002Flever\u002Fblob\u002Fmain\u002Fskills\u002Flever\u002FSKILL.md",[],[35,87,90,133,115,153,154,155,131,156],"setup","help","land","personal"]