Audit Agent Instructions for Accuracy and Risk
Stale commands, broken paths, and vague rules make agent instruction files unreliable. This skill combines local checks with evidence-based review to propose shorter, accurate instructions.
将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。
Review the Skillstore skill "agents-md-audit" from https://skillstore.io/skills/agentkitco-agents-md-audit.md and its manifest at https://skillstore.io/api/skills/agentkitco-agents-md-audit/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.你的智能体仍应展示其计划，并按安全策略请求必要确认。
当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。
正在使用“agents-md-audit”。 An instruction file refers to a missing relative directory and repeats a long rule.
预期结果:
正在使用“agents-md-audit”。 An instruction file includes a possible credential and a deployment command.
预期结果:
正在使用“agents-md-audit”。 Two instruction files require different package managers.
预期结果:
All 13 static findings are false positives involving Markdown, Python strings, comments, detection patterns, or dictionary operations. The checker reads local files without executing extracted commands or sending data over a network. The workflow permits safe local checks, excludes deployment and destructive commands, and requires approval before editing; no additional semantic security findings were identified.
分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。
https://skillstore.io/skills/agentkitco-agents-md-audit/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report[](https://skillstore.io/skills/agentkitco-agents-md-audit?utm_source=security_passport_badge)<a href="https://skillstore.io/skills/agentkitco-agents-md-audit?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/agentkitco-agents-md-audit/security.svg" alt="Skillstore security assessment" loading="lazy"></a><iframe src="https://skillstore.io/embed/skills/agentkitco-agents-md-audit.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>agentkitco. (2026). agents-md-audit security audit report (audit version 1) [Author version 1.0]. Skillstore. https://skillstore.io/skills/agentkitco-agents-md-audit/audits/1@techreport{agentkitco-agentkitco-agents-md-audit-2026,
author = {agentkitco},
title = {agents-md-audit security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/agentkitco-agents-md-audit/audits/1},
note = {Author version 1.0}
}cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agents-md-audit security audit report (audit version 1)"
version: "1.0"
type: report
authors:
- name: "agentkitco"
date-released: "2026-10-17"
url: "https://skillstore.io/skills/agentkitco-agents-md-audit/audits/1"
identifiers:
- type: other
value: "skillstore:agentkitco-agents-md-audit:audit:1"
description: "Skillstore immutable audit report identifier"
Review AGENTS.md after repository changes and identify outdated commands, missing paths, and instruction lines that need correction.
Compare Claude, Cursor, and Copilot instructions for conflicting rules, duplication, and guidance that belongs in scoped files.
Identify possible secrets, unresolved placeholders, and unverified commands before distributing repository instructions.
Audit AGENTS.md for missing paths, vague rules, duplicates, and placeholders. Report evidence by line and wait for approval before editing.
Audit CLAUDE.md against repository scripts and targets. Do not execute repository commands; mark runtime behavior as unverified and propose corrections supported by metadata.
Find agent instruction files and identify their readers. Compare rules for contradictions, redundant guidance, and scope problems, citing file names and line numbers.
Audit all instruction files using the checker and manual review. Separate required fixes, trimming suggestions, contradictions, and unverified commands.
作者
agentkitco许可证
MIT
作者版本
v1.0
Skillstore 修订
r1
版本提示
作者声明的版本不符合 SemVer 规范。
引用
743429f75b17204d47661a86b577a575ad67517f
维护新鲜度
2026/10/17
使用量
0 次下载 · 0 次浏览
文件结构
Debug Failures With Reproducible Evidence
Bug fixes based on guesses can miss the cause and introduce regressions. This skill guides reproduction, hypothesis testing, focused fixes, and verification with recorded evidence.
Verify Tests Ran Before Reporting Completion
Completion claims can hide skipped tests, stale results, or checks that never exercised a change. This skill guides fresh verification and reports evidence for each claim.
Write Pull Request Descriptions From Real Changes
Reviewers need clear explanations of changes, testing, and risk. This skill turns Git diffs and commits into structured pull request descriptions grounded in available evidence.
Upgrade Dependencies With Release Note Verification
Dependency upgrades can introduce breaking changes and obscure existing failures. This skill guides focused upgrades through release notes, baseline checks, targeted edits, and verification reports.
Preserve Progress Across Agent Sessions
Long sessions and agent changes can lose important project context. This skill creates a concise handoff with verified state, decisions, and ordered next steps.
Review Code Changes for Security Risks
Security defects can hide in small changes to input handling, authentication, and configuration. This skill guides scoped code triage with evidence, realistic severity, and actionable fixes.
重构 Agent 指令文件
由 softaworks
大型 agent 指令文件会浪费上下文并隐藏重要规则。此技能会将它们拆分为一个精简的根文件和聚焦的链接文档。
创建 Token 高效的 LLM 文档
由 Aceek
面向 LLM 的文档常常因冗长解释和重复细节而浪费上下文。此技能可将项目知识转化为密集、结构化的助手文档。
创建技能文档模板
由 ANcpLua
当结构不清晰时，开始一个新的 AI 技能会很慢。此模板为作者提供了一个基础的 SKILL.md 大纲，涵盖概述、触发条件、流程规范、示例、要求和相关链接。
审计文档覆盖率
由 C0ntr0lledCha0s
团队在发布和交接之前需要清晰的证据来了解文档缺口。此技能使用结构化的审计工作流审查覆盖率、质量和优先级。
Maintain Accurate Agent Context
由 drix10
Stale context files mislead agents and increase maintenance risk. Gardener verifies documentation against code, repairs references, and keeps risk baselines controlled.
扫描并整合项目指南
由 0xDarkMatter
项目指南通常分散在各种助手和 IDE 文件中。此技能可以查找这些文件、总结其中的共有规则，并提供一份整合后的 AGENTS.md。