All skills
kkkkhazix avatar

/neat-freak

@2b4a645
by Khazixkkkkhazix/khazix-skills21k stars
2,235

Knowledge and governance closeout: reconcile project docs, rule files (CLAUDE.md/AGENTS.md), authorized agent memory, and workspace residue with what the code and runtime actually do, so the next session or the next person starts from one current answer. Trigger when the user names "neat-freak", "洁癖", or "/neat" — and also on clear knowledge-closeout intent without the name: syncing or tidying project docs/rules/memory after development ("把文档和记忆整理一下", "收尾时把文档同步掉", "docs 和代码对不上了"), stale or conflicting CLAUDE.md/memory, a clean handoff to a teammate or a fresh session, or auditing whether workspace rules are actually followed. Do not trigger for pure coding/refactoring/debugging tasks, tidying data or prose (JSON, 周报, changelog announcements), or a bare "整理" with no project-knowledge context.

Use this Skill: https://skilld.dev/gh/kkkkhazix/khazix-skills/neat-freak

This session only. Nothing lands on disk.

referencesverification.md

≈1k tokens on demand. Your agent reads this file only when SKILL.md points to it.

证据层级与发布终态

风险决定证据深度

结论 最低证据
“文档链接有效” 项目自己的 doc-link/index check 或逐链接存在性
“规则已同源” realpath/readlink/import + 平台实际加载顺序
“代码实现是 X” 当前目标分支代码、schema、配置与相关测试
“PR 已完成” PR state=merged + merge commit;不能推导已部署
“已部署” deploy marker/release 指向目标 commit + 服务 active
“用户已看到新版本” canonical 用户 URL/API 的真实响应,必要时同时比 origin/cache
“可安全清场” merged + production contains change + knowledge receipt + lane clean + 无唯一未集成文件
“已获准清场” 完整结果已向用户汇报 + 用户在该汇报后明确确认可以清场 + 现场要求的确认凭证
“整个项目干净” 项目内所有适用事实面 verified;warning、pending 和 out-of-scope 单列

代码直觉、旧 memory、commit message 和 cache-buster URL 都只能当线索,不能单独证明生产终态。

真相矩阵

对每个发现至少记录:

topic: <事实主题>
authority: <当前权威来源>
code: verified-current | stale | n/a
runtime: verified-current | stale | unverified | n/a
docs: verified-current | stale | changed
rules: verified-current | stale | changed | n/a
memory: verified-current | stale | generated-read-only | changed | n/a
action: <做了什么或为什么没做>
verification: <命令、页面或门禁>

用户不需要看到完整矩阵,但最终摘要必须保留未闭合状态。

发布状态机

implemented
  -> locally verified
  -> pushed / PR opened
  -> CI + required backtest/visual review passed
  -> merged
  -> deployed
  -> live verified
  -> knowledge closed + receipt recorded
  -> full result reported while evidence is preserved
  -> user explicitly approved cleanup after the report
  -> workspace cleaned
  -> post-cleanup audit passed
  -> cleanup result appended

跳过的状态必须有项目规则允许的原因。失败停在哪一格,就按那一格汇报,不能用“基本完成”覆盖。

缓存和多表面产品

当用户可见内容经过 CDN、边缘缓存、搜索索引、异步 worker 或多客户端时,至少识别:

  • origin 是否为新内容;
  • canonical URL 是否仍为旧缓存;
  • API/页面/通知/RSS 是否共享同一数据出口;
  • deploy marker 是否在所有异步进程真正切换之后才写;
  • cache-buster 是否只是诊断,而非真实用户验收。

只验证其中一个表面时,在结论里明确限制范围。

清场前 gate

清场会销毁复盘和用户复核证据,因此顺序固定为:

  1. 验证目标工作已集成并上线;
  2. 同步 docs/rules/获准记忆;
  3. 记录项目要求的 knowledge closeout receipt;
  4. 预览待删除 worktree/branch/db/artifact;
  5. 检查 dirty 文件和 patch equivalence;
  6. 向用户完整汇报结果并保留上述现场;
  7. 等待用户在看完汇报后明确确认可以清场;
  8. 记录项目要求的用户确认凭证并执行授权的清理;
  9. 重新运行 workspace audit,补充汇报清场结果。

用户最初任务中的“收尾并清理”“做完删掉”等预授权不替代第 7 步;确认必须发生在完整汇报之后,因为用户要先看到结果才能判断是否需要保留现场继续复核。

目录名、分支年龄和 agent 会话是否关闭都不能证明可删除。

验证失败时

  • 同一失败第二次出现,停止盲重试,重新检查假设、环境和命令名。
  • 门禁要求机器可读 metadata 时,补正确留痕并触发新事件;不要用人工确认绕过可修复的格式问题。
  • 失败发生在生产写入前,明确说“尚未影响生产”;发生在切流后,先确认当前 active release 和回滚边界。
  • 任何未验证项保持 pending,不要为了摘要好看把它降格成 warning。

Source: SKILL.md on GitHub

No alerts16d3 checks · Risk SAFE
  • Gen Agent Trust Hub16d

    The skill 'neat-freak' is a project governance and documentation reconciliation tool designed to synchronize project documentation, agent rules, and memory with the actual state of the code. It includes strong safety guardrails, such as requiring explicit user confirmation for any destructive operations (like cleaning up temporary files or branches) and explicitly forbidding the exfiltration or storage of credentials discovered during audits. The skill uses a dedicated read-only Bash script to inventory project artifacts, and no malicious patterns or vulnerabilities were detected.

  • Socket16d

    No alerts

  • Snyk16d

    Risk: LOW · No issues

Signed by skilld at 2b4a645. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 7 hours ago.

Activeupdated 2 months ago
metadata
{
  "version": "3.0.0",
  "category": "knowledge-governance"
}
Other metadata
compatibility
Requires filesystem read access. Writes and destructive actions follow the active agent, workspace, and user authorization rules. Git and rg improve verification; scripts/audit-inventory.sh needs Bash — without it, do the equivalent checks manually. Works on any Agent Skills platform.

README badge

README badge for kkkkhazix/khazix-skills/neat-freak