跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
其他技能

ln-61-repository-publisher

Commits, pushes and remotely verifies authorized repository changes; does not create releases.

levnikolaevich · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add levnikolaevich/claude-code-skills --skill ln-61-repository-publisher

技能说明


name: ln-61-repository-publisher description: "Commits, pushes and remotely verifies authorized repository changes; does not create releases."

Repository Publisher

Goal: Publish only changes the user has authorized, then verify the result from the remote source.

Execution contract: The checklist defines completion. Track each item internally as PENDING, PROVEN with evidence, CLEARED with evidence its condition is absent, or UNPROVEN with a gap; reading, delegation, tool failure, a zero exit status, or a self-reported success is not proof; only the observed outcome is. Reconcile after each section. Before returning, resolve all PENDING, count only PROVEN and CLEARED, and apply verdict and approval rules to every gap. Preserve intent, scope, and existing authorization. Continue authorized work; ask only for consequential unresolved choices or required external approval. When no one can answer during the run, state the exact question and apply the skill's verdict for the remaining gap instead of waiting or guessing. Scale depth to material risk without skipping checks. Preserve dependency and safety order; otherwise choose an appropriate verification method. Accept equivalent user or repository evidence; no other skill, named artifact, or complete lifecycle is required. Preserve source requirement and decision IDs. Bind reused evidence to relevant source versions, dirty changes, configuration, and environment; invalidate only affected claims. Choose one primary workflow by the requested outcome, honoring explicit selection. Consume neighboring contracts and valid evidence without duplicating ownership or restarting a full audit. Split genuinely distinct deliverables by scope, not by mandatory skill invocation. Broad independent assessments belong to audit workflows; focused reviews may stay with the domain workflow but remain read-only. On continuation, reconcile task, authorization, current state, and unresolved evidence. For long work, return a compact continuation record or update an already authorized artifact; read-only skills do not persist it. Distinguish artifact readiness, verified behavior, and external-action authority. Prepare authorized work before required approval. If blocked by an instruction, cite its exact source and unresolved boundary; do not invent approval gates from caution. Before a new or changed technical decision, establish applicability and actual technology versions, then verify version-dependent assumptions against official documentation or specifications even when confident. Search and read primary pages for current external recommendations; search snippets are not evidence. Version-matched bundled documentation or still-valid prior evidence can suffice. Record the decision, applicable version, source and limitations. If evidence is unavailable, mark the affected conclusion UNVERIFIED and block only dependent work when essential. Prefer native mechanisms; add tools or abstractions only for an evidenced requirement. Reuse the existing project rule owner instead of creating parallel policy.

Tool Routing

NeedPreferred capabilityFallback
Scope, staging, history, and synchronizationNative Git CLIStop if equivalent Git evidence is unavailable
Repository validationProject-native commands and installed validatorsDocumented manual checks
Remote branch and CI stateAuthenticated hosting CLI or connectorGit remote evidence plus direct workflow URLs
Clean-source verificationTemporary clone and isolated host configuration when distribution surfaces changedRemote raw files plus BLOCKED for required install evidence
Static-site verificationDeployment workflow plus direct HTTP requestHosting API deployment state
Current marketplace behaviorOfficial host documentationMark assumptions and avoid destructive retries

Prefer compact Git output first, then open the full diff for files that will be staged. Never pipe commands in a way that hides the failing exit code.

Use hosting APIs for remote facts and Git for repository facts. Local distribution or deployment state cannot prove that a remote consumer update works.

Checklist

Establish scope and evidence

  • Confirm the user explicitly requested a commit and push and identify the intended branch.
  • Read repository instructions, release rules, and the current branch policy before mutation.
  • Check git status -sb, staged and unstaged diffs, untracked files, remotes, and recent commit style.
  • Identify unrelated user changes; do not stage them without explicit whole-worktree authorization.
  • Inspect deletions and generated files as carefully as edited text.
  • Check whether behavior, installation commands, catalogs, layout, or the public site require matching documentation updates.
  • Do not change versions, tags, or release metadata during an ordinary publication unless the request or repository policy explicitly includes them.
  • For marketplace edits, confirm every stable identifier is unchanged unless an intentional migration was approved.
  • When multiple host catalogs exist and repository policy requires parity, confirm they contain the same distribution units in the required order.
  • When metadata is duplicated across manifests or catalogs, confirm descriptions and source paths agree with the canonical source.

Validation Routing

  • Discover and run repository-native validation commands before generic checks.
  • For changed skills, run repository-required or host-native skill validators, or perform their documented manual fallback.
  • For changed plugins or packages, run repository-required validators, or perform their documented manual fallback.
  • Run every host-native strict validator whose distribution surface exists and changed or is required by repository policy.
  • Run only the catalog parity, manifest parsing, stale-reference, lo
阅读完整来源 ↗