Agent审查

Name: Agent审查
Author: tomevault-io

效率工具社区

作者仓库星标 0
作者更新于实时读取
作者仓库 skills-registry

流狐档案事实层信息：领域、Agent、信任分、运行环境、权限与网络

领域: 效率工具
兼容 Agent: Claude Code
Cursor
Cline
Codex
Windsurf
Gemini CLI

+20
信任分: 88 / 100 · 社区维护
作者 / 版本 / 许可: @tomevault-io · 未声明 license
Token 消耗评级: 低消耗
接入复杂程度: 需简单配置
是否需要外部 API Key: 不需要
兼容的系统: 未声明（默认跨平台）
底层运行要求: 无特殊要求
文件与系统权限: 只读
允许写入 / 修改
Shell 执行
网络行为: 允许外网请求
安装命令数: 26 条

档案由构建时根据 SKILL.md 与安装命令自动衍生，可能与作者实际意图存在差异。

需要注意： 未限定 allowed-tools，默认拥有全部工具权限。

输出预览 patch.preview

---
name: patch
description: Propose a code patch for a finding. Produces a unified diff against the current HEAD plus a shor…
category: 效率工具
runtime: 无特殊运行时
---

# patch 输出预览

## PART A: 任务判断
- 适用问题：日常重复事务、资料整理、邮件或工作流提效。
- 输入要求：目标材料、限制条件、期望输出和验收方式。
- 证据边界：围绕“Workspace / What to do / Refusing to patch”读取原文规则，不把推断写成作者承诺。

## PART B: 执行结果
- **01** 任务判断：确认你的需求是否属于日常重复事务、资料整理、邮件或工作流提效，并标出输入、限制和预期结果。
- **02** 执行计划：优先按“Workspace / What to do / Refusing to patch”拆成步骤，说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果：给出可复制的命令、文件改动、检查清单或内容草稿，并说明如何继续迭代。
- **04** 风险边界：结合 读取文件、写入/修改文件、执行终端命令、会按任务需要访问外部网络、通常不需要额外 API Key 给出执行前确认项。

## Running Rules
- 读取文件、写入/修改文件、执行终端命令；会按任务需要访问外部网络；通常不需要额外 API Key。
- 先小样例验证，再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。

SKILL.md SKILL.md

---
name: patch
description: Propose a code patch for a finding. Produces a unified diff against the current HEAD plus a shor…
category: 效率工具
source: tomevault-io/skills-registry
---

# patch

## 什么时候使用
- 用于审阅代码、文档或方案并给出可执行反馈 适合处理重复事务、信息整理、邮件和日常工作流提效，核心价值是把输入、判断、执行、验证和交付边界固定下来，避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步骤；通常不需要额外…
- 面向日常重复事务、资料整理、邮件或工作流提效，优先处理能明确输入、步骤和验收标准的工作。

## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。

## 执行规则
- 围绕「Workspace / What to do / Refusing to patch」组织步骤，不把推断写成作者事实。
- 读取文件、写入/修改文件、执行终端命令；会按任务需要访问外部网络；通常不需要额外 API Key。
- 先跑小样例，确认结果可检查后再扩大任务范围。

## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown，不编造命令、平台或依赖。

证据边界与执行链路

作者原文负责流程事实；仓库文件负责来源和命令；流狐只补充适用场景、限制和质量判断。

skill "patch" {
  输入层     -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
  上下文层   -> Workspace / What to do / Refusing to patch
  规则层     -> SKILL.md 触发条件 / 执行顺序 / 输出格式
  运行层     -> 无特殊运行时 | 读取文件、写入/修改文件、执行终端命令 | 会按任务需要访问外部网络
  安全层     -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
  输出层     -> 可复制结果 + 检查清单 + 下一步迭代
}

解读按原文结构重写，命令、链接、术语均保留；右侧可核对作者原始 SKILL.md

先判断是否适合

核心用途：用于审阅代码、文档或方案并给出可执行反馈适合处理重复事务、信息整理、邮件和日常工作流提效，核心价值是把输入、判断、执行、验证和交付边界固定下来，避免 Agent 泛泛回答。把任务拆成可执行、可检查、可继续迭代的步骤；通常不需要额外 API Key，适合…
适用任务：更适合重复事务、信息整理、邮件和日常工作流提效，尤其是输入、步骤、验收口径可以反复复用的场景。
不宜硬套：如果任务只有一次性问答、来源材料很少，或需要作者没有说明的命令和外部服务，应先降级为普通对话。

作者设计意图

触发依据：使用时机、上下文收集、输出格式和失败处理共同决定它是否适合当前任务。
来源边界：作者原文负责事实，仓库文件负责命令和文件证据；Fluxly 只补充用途、限制和使用判断。

作者的方法与取舍

执行主线：任务围绕「Workspace / What to do / Refusing to patch」组织，把触发条件、必要上下文、执行规则和交付标准连成一条可复核的链路。
设计重点：重点不在重复教程，而在解释作者为什么规定这些步骤、哪些条件不可省略，以及结果偏离时应回到哪一层检查。
使用判断：先确认任务与原文边界一致，再决定是否启用；来源没有说明的命令、平台和依赖保持未知，不由 Fluxly 补写。

边界和复核

依赖条件：通常不需要额外 API Key，适合先用小任务验证触发是否稳定。
权限提醒：会涉及 read / write / shell-exec 权限，执行前要说明可读写范围、命令边界和回退方式。
质量标准：结果应包含明确产物、必要证据和下一步动作；如果只是泛泛解释，需要补充输入和验收条件后重跑。

patch

Propose a minimal code patch that fixes a confirmed finding. You are not shipping the fix; you are handing the analyst a starting diff and explaining what it does. The analyst reviews, edits if needed, and opens a PR by hand.

Workspace

./src — the repository at its current HEAD, on the default branch, writable
./context.json — has scrutineer.api_base, scrutineer.token, scrutineer.repository_id, scrutineer.finding_id (required; this skill only makes sense finding-scoped)
./report.json — write the patch + rationale here
./schema.json — shape of report.json

What to do

Read ./context.json. If scrutineer.finding_id is missing, write {"error": "no finding_id in context.json; patch is finding-scoped"} to report.json and exit 0.
Fetch the finding: GET {api_base}/findings/{finding_id} with Authorization: Bearer {token}. Read location, cwe, trace, boundary, validation, rating. These five together tell you where the sink is, what the vulnerable input flow looks like, and what dangerous behaviour you need to stop.
Inside ./src, edit files to fix the finding. Constraints:
- Minimal. Change only what the fix requires. Do not refactor surrounding code, rename variables, reformat unrelated lines, or upgrade dependencies unless the fix inherently requires it.
- In place. Fix the sink where it lives. If the finding's location is pkg/foo/bar.go:42, that is where the patch should land (or at the nearest layer where a guard is sensible — e.g. the input validator that feeds the sink).
- Consistent. Match the existing code style and idioms. If the codebase uses a specific sanitiser, validator, or helper for similar cases, reuse it. Do not introduce a new helper module for a one-off fix.
- Safe. The patch must not break the reproduction's documented legitimate behaviour — only block the dangerous path. If you cannot tell where the dangerous path diverges from legitimate use, stop and write an inconclusive report (see below).
- Include a test when practical. If the repo has a test suite that covers the vulnerable code path, add a regression test that would fail without your patch. If the repo has no tests, or the sink is in a place that is hard to cover, skip this and say why in rationale.
Once you have a working tree edit, generate a unified diff against HEAD:
```
cd src
git add -N .
git diff HEAD -- . > /tmp/patch.diff
```
Read /tmp/patch.diff and put its contents into report.json under the patch field. Do not commit; the diff is the artefact. If the diff is empty, something went wrong — do not write an empty patch. Write {"error": "patch produced no diff"} and exit 0.
POST a finding note summarising the patch: POST {api_base}/findings/{finding_id}/notes with:
```
{
  "body": "Proposed patch in scan #{scan_id}.\n\nFiles changed: ...\n\n{short rationale}\n\nApply with: `git apply` the diff from the scan report.",
  "by": "patch"
}
```
The note lives on the finding page; the full diff lives in report.json and is viewable on the scan page.
Do not PATCH any editable fields on the finding. Specifically:
- Do not set fix_commit — that field means a shipped upstream fix, not a proposal. The analyst sets it after their PR merges.
- Do not set fix_version — same reason.
- Do not touch status. Lifecycle transitions belong to the analyst.

Write ./report.json:

{
  "patch": "diff --git a/pkg/foo/bar.go b/pkg/foo/bar.go\n...",
  "rationale": "Short prose — two or three sentences. What the guard is, why it blocks the trace, what legitimate input it still lets through.",
  "files_changed": ["pkg/foo/bar.go", "pkg/foo/bar_test.go"],
  "base_commit": "<HEAD sha from ./src>",
  "tests_added": true,
  "notes": "Optional: anything the analyst should know — a second sink you spotted but didn't patch, a style choice you weren't sure of, a test you couldn't write."
}

base_commit is the HEAD sha the diff applies to. The analyst needs this to git am or git apply cleanly — if they rebased since the scan, they know the patch may not apply and can regenerate.

Refusing to patch

Write {"error": "...", "rationale": "..."} and exit 0 in any of these cases — do not ship a bad patch:

The finding prose is too thin (empty Trace, empty Validation). You need both to know where the sink is and what behaviour to stop.
The fix is architectural (e.g. "rewrite this whole module to not shell out") rather than localisable. A patch skill proposes a surgical fix; larger changes are an issue comment for the maintainer, not a diff.
The codebase is in a language or framework you cannot confidently edit without risking regressions. It is better to say so than to produce a plausible-looking but wrong patch.
The finding has already been fixed upstream. Check git log -- {location} — if a recent commit looks like it addressed the sink, surface the SHA in notes and refuse to duplicate.

Constraints

Do not push. Do not commit. Do not open a PR. The scrutineer workspace is ephemeral and isolated; your diff is the only thing that survives the scan.
Do not add dependencies unless the vulnerability genuinely requires one (e.g. a sanitiser library the codebase already uses elsewhere). New top-level deps in a patch almost always mean the fix is in the wrong place.
Do not edit the lockfile, go.sum, Gemfile.lock, package-lock.json, Cargo.lock, etc. unless you also changed the manifest that owns it. Stray lockfile churn makes diffs hard to review.
Do not touch files outside what the fix requires. CI config, docs unrelated to the fix, README — leave alone.

Source: alexandre-daubois/scrutineer — distributed by TomeVault.

安装命令

一键复制安装选择包管理器,点击右侧复制,粘贴到终端回车即可

npx skills add tomevault-io/skills-registry/alexandre-daubois--scrutineer--patch -y

查看通用安装教程

查看 Skill 原文

兼容工具

Claude Code
Cursor
Cline
Codex
Windsurf
Gemini
OpenCode
Roo Code
AAide
+Augment
GitHub Copilot
OpenClaw
KKiro
Junie
WWarp
Goose
Trae
Kilo Code
OpenHands
Amp
Zencoder
CContinue
AIAider
CodeBuddy
CodeGeeX
Replit Agent

安全审计

88 / 100

社区

未声明 license/许可。
未声明 version。
未限定 allowed-tools，默认拥有全部工具权限。

元信息

来源: github
tomevault-io/skills-registry
English

@tomevault-io 的其他技能

分类与标签

分类效率工具