[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7D10E291WGS1MWPZPW1BPR7":3,"skill-same:01M7D10E291WGS1MWPZPW1BPR7":61},{"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":6,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":31,"install":32,"resources":44},"01M7D10E291WGS1MWPZPW1BPR7","fix-pr",null,"Diagnose and repair failed GitHub Actions workflows for pull requests without weakening required checks, then verify the repaired commit.","grove\u002Fpromise-to-proof","grove","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F326795?v=4","01M7D0ZFW80DSGZY7G3SS3YGWH","Geir O. Grønmo","01M7D0ZFXA8ZWXE7BKH69GKRCH",4,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ffix-pr\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: fix-pr\ndescription: Diagnose and repair failed GitHub Actions workflows for pull requests without weakening required checks, then verify the repaired commit.\ndisable-model-invocation: true\n---\n\n`\u002Ffix-pr` restores a pull request's intended workflow outcome. It identifies the\nearliest causal failure, classifies its cause, applies the smallest durable\nrepair, and verifies the repaired commit against every required check.\n\n## Local work and durable records\n\nUse the [filesystem protocol](references\u002Facceptance-contract-protocol.md) and\n`python3 \u003Cskill-dir>\u002Fscripts\u002Fp2p_filesystem.py --repo \u003Croot> resolve .p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md`\nto resolve paths. Save reports with `save .p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md \u003Creport-name> --from \u003Cfile>`\nto retain history before replacement.\n\nWhen a canonical `.p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md` applies, discover its saved candidate and\nreports in `.p2p\u002Fwork\u002F\u003Cslug>\u002F`. Save diagnosis, repair outcome, retained safe\nevidence, and exact old\u002Fnew candidate identities to\n`.p2p\u002Fwork\u002F\u003Cslug>\u002Fci-repair.md`, preserving prior records under the protocol\nretention rule. Put disposable logs in `.p2p\u002Ftmp\u002F`; retain needed evidence before\nclaiming a handoff. Redact secrets, and use the protocol durable reference and\nchecksum record for evidence unsuitable for Git. Local report saving alone does\nnot grant Git authority; the explicit `\u002Ffix-pr` invocation authority below remains\nseparate. Candidate changes require fresh review\u002Fproof; never rewrite old verdicts.\n\n## Invocation\n\nAccept one repository target:\n\n- `\u002Ffix-pr #123`\n- `\u002Ffix-pr https:\u002F\u002Fgithub.com\u002Facme\u002Fproject\u002Fpull\u002F123`\n- `\u002Ffix-pr \u003Cworkflow-run-url>`\n\nInfer the relevant workflow from a pull request or workflow run whenever the\ntarget makes that possible.\n\n## Guardrails\n\nBefore editing, inspect the branch and worktree. Continue only when the\ncurrent branch and commit identify the target PR head or workflow commit, and\nthe worktree has no unrelated changes. Stop with `NOT FIXED` when the target\ncannot be established, credentials or permissions are unavailable, or the\nbranch state makes the repair ambiguous.\n\nCommit only the scoped repair, push only the target PR head branch, and never\nforce-push.\n\nBind diagnosis and verification to the exact target commit SHA. If the PR\nchanges during the repair, stop and re-establish the target rather than\nsilently chasing the newer commit.\n\nFor fork or otherwise untrusted PRs, do not expose secrets, approve privileged\nworkflow execution, broaden permissions, or weaken isolation. Stop with\n`NOT FIXED` when the required verification depends on an unavailable safe\nexecution path.\n\nPreserve the verification boundary that exposed the failure:\n\n- Keep every required job, test, assertion, lint\u002Ftype check, coverage\n  threshold, CI matrix entry, and failure condition. Replace an incorrect or\n  obsolete check only under the evidence rule in `Forbidden repairs`.\n- For a workflow-only repair, require the same quality checks to run with the\n  same failure semantics.\n- Add or strengthen a focused regression test for a product or test defect\n  when the repository has a suitable seam. Use an independent behavioral\n  assertion, not a copy of the implementation's calculation.\n- Keep fixes scoped to the failure. A rerun is evidence, not a repair.\n\n## Workflow\n\n### 1. Establish the intended operation\n\nDetermine what the failed workflow was meant to guarantee, such as validating\nthe PR branch before merge or running required checks for a specific commit.\nDefine success from that guarantee.\n\nRead the [acceptance contract protocol](references\u002Facceptance-contract-protocol.md)\nand the linked ticket and acceptance contract when they define the behavior under\nrepair. Preserve the revision, requirement IDs, promised results, boundaries,\nagreed seams, and independent oracles. If that contract is unavailable or\nambiguous, continue only with repairs whose correctness does not depend on\nguessing the missing requirement.\nReport the gap; a CI failure does not authorize changing the ticket's promises.\n\n### 2. Diagnose before editing\n\nInspect the workflow run, failed jobs and steps, logs, workflow YAML, relevant\nchanges, and the target PR's branch-protection or ruleset-required checks.\nStart at the earliest causal failure, then look for independent failures in\nparallel jobs. Record the workflow, job, step, command, and observed error.\n\nChoose local reproduction, a pushed GitHub Actions run, or both by expected\nevidence per unit of clock time, context cost, and compute cost. Use local\nchecks when they isolate the cause quickly; prefer the authoritative remote\nworkflow when it is environment-specific or expensive to reproduce locally.\nUse existing artifacts and focused static inspection when they answer the\nquestion; avoid equivalent duplicate runs.\n\nClassify the primary cause before changing anything:\n\n- `PRODUCT`: branch code, test target, or generated output is wrong.\n- `WORKFLOW`: GitHub Actions configuration, permissions, or job wiring is\n  wrong.\n- `EXTERNAL`: dependency, runner, credential, or service is unavailable.\n\nFor `EXTERNAL`, do not change product or workflow code to compensate for the\nunavailable capability. Resolve the external condition only when it is\nauthorized and independently verifiable; otherwise return `NOT FIXED`.\n\nTreat flakiness as a confidence modifier. A rerun without a relevant change\nends as `NOT FIXED — WORKFLOW GREEN — FLAKE NOT RESOLVED` until nondeterminism\nis understood.\n\n### 3. Repair and verify\n\nApply the smallest complete repair at the layer that owns the failure. Keep\nrequired checks and their failure semantics, subject to the evidence rule in\n`Forbidden repairs`. For product and test\ndefects, add or strengthen the focused regression evidence before declaring\nsuccess.\n\nMap product or test repairs to the affected requirements and their observable\nresults. Record replacement evidence and why it still checks the same promise.\nKeep unrelated acceptance gaps visible without expanding this CI repair into a\nfull implementation or proof run.\n\nRun the cheapest high-signal verification first, whether local or remote. Use\nthe authoritative GitHub workflow as the primary verification when it is the\nmore efficient evidence source; run broader local checks only when their\nadditional evidence justifies the cost.\nVerify every required check reported by the target PR's branch protection or\nruleset, not only the original failed job. For a workflow-run target, use its\nassociated PR when available; otherwise verify every check represented by the\nrun and state that branch-protection requirements could not be established.\nCount a required check only when it succeeds for the exact repaired SHA.\nPending, queued, cancelled, timed-out, skipped, missing, or stale results are\nnot successful verification. Use the environment's waiting mechanism when one is\navailable; do not invent polling intervals, wait budgets, or repair-attempt caps.\nWithout a waiting mechanism, return `NOT FIXED — VERIFICATION PENDING` with the\nrun URLs. Stop when the cause is uncertain, a required capability is unavailable,\nthe repair would weaken quality, or the repair introduces an unexplained quality\nregression. Confirm that the repaired commit is pushed and all required signals\nremain measurable.\n\n## Forbidden repairs\n\nKeep the required quality signal measurable. Do not delete or skip tests,\ndisable lint rules, lower coverage, make checks optional, swallow failures,\nadd unconditional `continue-on-error`, add arbitrary retries, or comment out\nworkflow steps. A check may change only when the source contract and evidence\nprove it incorrect or obsolete. The replacement must provide stronger valid\nevidence for the same promised behavior. Record the reason and affected\nrequirement IDs; changing a check does not authorize changing the promise.\n\nDo not redesign workflows, clean up unrelated code, or use a green rerun to\nmask an unresolved failure.\n\n## Completion\n\nReturn one of these outcomes:\n\n- `FIXED`: the cause is understood, a durable repair and focused verification\n  pass, required quality signals are preserved or strengthened, the repaired\n  commit is pushed, and all required checks for the target are green.\n- `NOT FIXED`: the failure remains unresolved, required capability is\n  unavailable, the worktree is unsafe, the repair would weaken quality, or\n  only an unexplained rerun passed.\n\n`FIXED` means the workflow is repaired. It does not mean every ticket requirement\nis `proven`.\n\nFor either outcome, report the intended operation, failure classification,\nearliest causal error, files changed, verification performed, required-check\nstatus, and the remaining reason when `NOT FIXED`.\nInclude the repaired SHA, affected requirement IDs, replaced evidence, and any\nacceptance gaps. Record the preserved contract revision when one exists.\nAny changed candidate leaves prior proof tied to the old candidate. Changes to\nproduct behavior, acceptance evidence, or relevant tests require fresh `\u002Fprove`\nagainst every requirement before acceptance. Green CI cannot refresh proof.\nIdentify stale proof and review results, and hand off to `\u002Fprove` and the\nrepository's review process when those steps are requested or required.\n\nInvoking `\u002Ffix-pr` authorizes scoped inspection, edits, commit, push, and\nworkflow reruns for the target PR or workflow run. It does not authorize\nforce-pushes, changes outside the target branch, or workarounds that weaken\nverification.\n\nEnd with `Next steps:` and a numbered list (`1.`, `2.`, ...) of applicable\nactions in order, so each can be referenced by number. For `FIXED`, refresh stale\nreports with the applicable invocation below; when both match, give\n`\u002Fmerge-readiness \u003CPR URL>` with the saved report references.\n\n```text\n\u002Fprove \u003Csaved contract>; candidate \u003Crepaired SHA>\n\u002Freview-implementation \u003Crepaired candidate> against \u003Ccomparison base>\n```\n\nFor `NOT FIXED`, give the exact failed check command from the workflow and its\nobserved failure, or name the specific missing input or human decision. Do not\nclaim acceptance or merge authority.\n","`\u002Ffix-pr` restores a pull request's intended workflow outcome. It identifies the\nearliest causal failure, classifies its cause, applies the smallest durable\nrepair, and verifies the repaired commit against every required check.\n\n## Local work and durable records\n\nUse the [filesystem protocol](references\u002Facceptance-contract-protocol.md) and\n`python3 \u003Cskill-dir>\u002Fscripts\u002Fp2p_filesystem.py --repo \u003Croot> resolve .p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md`\nto resolve paths. Save reports with `save .p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md \u003Creport-name> --from \u003Cfile>`\nto retain history before replacement.\n\nWhen a canonical `.p2p\u002Fwork\u002F\u003Cslug>\u002Fcontract.md` applies, discover its saved candidate and\nreports in `.p2p\u002Fwork\u002F\u003Cslug>\u002F`. Save diagnosis, repair outcome, retained safe\nevidence, and exact old\u002Fnew candidate identities to\n`.p2p\u002Fwork\u002F\u003Cslug>\u002Fci-repair.md`, preserving prior records under the protocol\nretention rule. Put disposable logs in `.p2p\u002Ftmp\u002F`; retain needed evidence before\nclaiming a handoff. Redact secrets, and use the protocol durable reference and\nchecksum record for evidence unsuitable for Git. Local report saving alone does\nnot grant Git authority; the explicit `\u002Ffix-pr` invocation authority below remains\nseparate. Candidate changes require fresh review\u002Fproof; never rewrite old verdicts.\n\n## Invocation\n\nAccept one repository target:\n\n- `\u002Ffix-pr #123`\n- `\u002Ffix-pr https:\u002F\u002Fgithub.com\u002Facme\u002Fproject\u002Fpull\u002F123`\n- `\u002Ffix-pr \u003Cworkflow-run-url>`\n\nInfer the relevant workflow from a pull request or workflow run whenever the\ntarget makes that possible.\n\n## Guardrails\n\nBefore editing, inspect the branch and worktree. Continue only when the\ncurrent branch and commit identify the target PR head or workflow commit, and\nthe worktree has no unrelated changes. Stop with `NOT FIXED` when the target\ncannot be established, credentials or permissions are unavailable, or the\nbranch state makes the repair ambiguous.\n\nCommit only the scoped repair, push only the target PR head branch, and never\nforce-push.\n\nBind diagnosis and verification to the exact target commit SHA. If the PR\nchanges during the repair, stop and re-establish the target rather than\nsilently chasing the newer commit.\n\nFor fork or otherwise untrusted PRs, do not expose secrets, approve privileged\nworkflow execution, broaden permissions, or weaken isolation. Stop with\n`NOT FIXED` when the required verification depends on an unavailable safe\nexecution path.\n\nPreserve the verification boundary that exposed the failure:\n\n- Keep every required job, test, assertion, lint\u002Ftype check, coverage\n  threshold, CI matrix entry, and failure condition. Replace an incorrect or\n  obsolete check only under the evidence rule in `Forbidden repairs`.\n- For a workflow-only repair, require the same quality checks to run with the\n  same failure semantics.\n- Add or strengthen a focused regression test for a product or test defect\n  when the repository has a suitable seam. Use an independent behavioral\n  assertion, not a copy of the implementation's calculation.\n- Keep fixes scoped to the failure. A rerun is evidence, not a repair.\n\n## Workflow\n\n### 1. Establish the intended operation\n\nDetermine what the failed workflow was meant to guarantee, such as validating\nthe PR branch before merge or running required checks for a specific commit.\nDefine success from that guarantee.\n\nRead the [acceptance contract protocol](references\u002Facceptance-contract-protocol.md)\nand the linked ticket and acceptance contract when they define the behavior under\nrepair. Preserve the revision, requirement IDs, promised results, boundaries,\nagreed seams, and independent oracles. If that contract is unavailable or\nambiguous, continue only with repairs whose correctness does not depend on\nguessing the missing requirement.\nReport the gap; a CI failure does not authorize changing the ticket's promises.\n\n### 2. Diagnose before editing\n\nInspect the workflow run, failed jobs and steps, logs, workflow YAML, relevant\nchanges, and the target PR's branch-protection or ruleset-required checks.\nStart at the earliest causal failure, then look for independent failures in\nparallel jobs. Record the workflow, job, step, command, and observed error.\n\nChoose local reproduction, a pushed GitHub Actions run, or both by expected\nevidence per unit of clock time, context cost, and compute cost. Use local\nchecks when they isolate the cause quickly; prefer the authoritative remote\nworkflow when it is environment-specific or expensive to reproduce locally.\nUse existing artifacts and focused static inspection when they answer the\nquestion; avoid equivalent duplicate runs.\n\nClassify the primary cause before changing anything:\n\n- `PRODUCT`: branch code, test target, or generated output is wrong.\n- `WORKFLOW`: GitHub Actions configuration, permissions, or job wiring is\n  wrong.\n- `EXTERNAL`: dependency, runner, credential, or service is unavailable.\n\nFor `EXTERNAL`, do not change product or workflow code to compensate for the\nunavailable capability. Resolve the external condition only when it is\nauthorized and independently verifiable; otherwise return `NOT FIXED`.\n\nTreat flakiness as a confidence modifier. A rerun without a relevant change\nends as `NOT FIXED — WORKFLOW GREEN — FLAKE NOT RESOLVED` until nondeterminism\nis understood.\n\n### 3. Repair and verify\n\nApply the smallest complete repair at the layer that owns the failure. Keep\nrequired checks and their failure semantics, subject to the evidence rule in\n`Forbidden repairs`. For product and test\ndefects, add or strengthen the focused regression evidence before declaring\nsuccess.\n\nMap product or test repairs to the affected requirements and their observable\nresults. Record replacement evidence and why it still checks the same promise.\nKeep unrelated acceptance gaps visible without expanding this CI repair into a\nfull implementation or proof run.\n\nRun the cheapest high-signal verification first, whether local or remote. Use\nthe authoritative GitHub workflow as the primary verification when it is the\nmore efficient evidence source; run broader local checks only when their\nadditional evidence justifies the cost.\nVerify every required check reported by the target PR's branch protection or\nruleset, not only the original failed job. For a workflow-run target, use its\nassociated PR when available; otherwise verify every check represented by the\nrun and state that branch-protection requirements could not be established.\nCount a required check only when it succeeds for the exact repaired SHA.\nPending, queued, cancelled, timed-out, skipped, missing, or stale results are\nnot successful verification. Use the environment's waiting mechanism when one is\navailable; do not invent polling intervals, wait budgets, or repair-attempt caps.\nWithout a waiting mechanism, return `NOT FIXED — VERIFICATION PENDING` with the\nrun URLs. Stop when the cause is uncertain, a required capability is unavailable,\nthe repair would weaken quality, or the repair introduces an unexplained quality\nregression. Confirm that the repaired commit is pushed and all required signals\nremain measurable.\n\n## Forbidden repairs\n\nKeep the required quality signal measurable. Do not delete or skip tests,\ndisable lint rules, lower coverage, make checks optional, swallow failures,\nadd unconditional `continue-on-error`, add arbitrary retries, or comment out\nworkflow steps. A check may change only when the source contract and evidence\nprove it incorrect or obsolete. The replacement must provide stronger valid\nevidence for the same promised behavior. Record the reason and affected\nrequirement IDs; changing a check does not authorize changing the promise.\n\nDo not redesign workflows, clean up unrelated code, or use a green rerun to\nmask an unresolved failure.\n\n## Completion\n\nReturn one of these outcomes:\n\n- `FIXED`: the cause is understood, a durable repair and focused verification\n  pass, required quality signals are preserved or strengthened, the repaired\n  commit is pushed, and all required checks for the target are green.\n- `NOT FIXED`: the failure remains unresolved, required capability is\n  unavailable, the worktree is unsafe, the repair would weaken quality, or\n  only an unexplained rerun passed.\n\n`FIXED` means the workflow is repaired. It does not mean every ticket requirement\nis `proven`.\n\nFor either outcome, report the intended operation, failure classification,\nearliest causal error, files changed, verification performed, required-check\nstatus, and the remaining reason when `NOT FIXED`.\nInclude the repaired SHA, affected requirement IDs, replaced evidence, and any\nacceptance gaps. Record the preserved contract revision when one exists.\nAny changed candidate leaves prior proof tied to the old candidate. Changes to\nproduct behavior, acceptance evidence, or relevant tests require fresh `\u002Fprove`\nagainst every requirement before acceptance. Green CI cannot refresh proof.\nIdentify stale proof and review results, and hand off to `\u002Fprove` and the\nrepository's review process when those steps are requested or required.\n\nInvoking `\u002Ffix-pr` authorizes scoped inspection, edits, commit, push, and\nworkflow reruns for the target PR or workflow run. It does not authorize\nforce-pushes, changes outside the target branch, or workarounds that weaken\nverification.\n\nEnd with `Next steps:` and a numbered list (`1.`, `2.`, ...) of applicable\nactions in order, so each can be referenced by number. For `FIXED`, refresh stale\nreports with the applicable invocation below; when both match, give\n`\u002Fmerge-readiness \u003CPR URL>` with the saved report references.\n\n```text\n\u002Fprove \u003Csaved contract>; candidate \u003Crepaired SHA>\n\u002Freview-implementation \u003Crepaired candidate> against \u003Ccomparison base>\n```\n\nFor `NOT FIXED`, give the exact failed check command from the workflow and its\nobserved failure, or name the specific missing input or human decision. Do not\nclaim acceptance or merge authority.\n",{"name":5,"description":7,"disable-model-invocation":24},true,"main","17f3deb982a8224f71b12bda74be6b24d86571e4","skills\u002Fproductivity\u002Ffix-pr",1,"https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof","https:\u002F\u002Fgithub.com\u002Fgrove","2026-10-12 16:06:12",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":24},[34,37,40],{"tool":35,"command":36},"npx","npx skills add grove\u002Fpromise-to-proof --skill fix-pr",{"tool":38,"command":39},"bunx","bunx skills add grove\u002Fpromise-to-proof --skill fix-pr",{"tool":41,"command":42},"pnpm","pnpm dlx skills add grove\u002Fpromise-to-proof --skill fix-pr","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ffix-pr\u002FSKILL.md",[45,51,55],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"agents\u002Fopenai.yaml","REFERENCE","参考文档",154,"https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ffix-pr\u002Fagents\u002Fopenai.yaml",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Facceptance-contract-protocol.md",48,"https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ffix-pr\u002Freferences\u002Facceptance-contract-protocol.md",{"path":56,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":59,"sourceHtmlUrl":60},"scripts\u002Fp2p_filesystem.py","SCRIPT","脚本",45,"https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ffix-pr\u002Fscripts\u002Fp2p_filesystem.py",[62,69,76,83,90,97],{"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},"01M7D123RRDGW8JKZZ635Q22CP","triage-issue","Assess an existing issue, recommend its next action and triage label, and apply only explicitly approved issue changes.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Ftriage-issue\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},"01M7D11YTHASCMV5C642SNXHGR","slice-contract","Divide a parent work item into complete local child work items with contribution mappings; optionally mirror an approved breakdown to a tracker.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Fslice-contract\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},"01M7D11W0WJEH1QVCVZAMZM6YW","setup-promise-to-proof","Set up ignored .p2p workflow storage, with optional tracker, triage labels, and domain-doc pointers.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Fsetup-promise-to-proof\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},"01M7D11Q1Z9RATB4G4VFTNG2GN","review-implementation","Review a fixed implementation against its acceptance contract, scope, and engineering obligations without editing it or declaring acceptance.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Freview-implementation\u002FSKILL.md",[],[],{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":19},"01M7D11KYBEARKFX1FRB1B18FT","retrospect","Evaluate a proven delivery after acceptance and propose evidence-backed advisory learnings for future work.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Fretrospect\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M7D11G9PE8TP0T0XCV0157H1","replan-backlog","Reconcile affected open issues with shipped code and completed PRs, proposing the smallest justified backlog changes.","https:\u002F\u002Fgithub.com\u002Fgrove\u002Fpromise-to-proof\u002Fblob\u002Fmain\u002Fskills\u002Fproductivity\u002Freplan-backlog\u002FSKILL.md",[],[]]