Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add SimpleEve/eo-skills --skill eo-change-reviewgit clone --depth 1 https://github.com/SimpleEve/eo-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/simpleeve/eo-skills/eo-change-review)<a href="https://agentmods.dev/skills/simpleeve/eo-skills/eo-change-review"><img src="https://agentmods.dev/badge/skills/simpleeve/eo-skills/eo-change-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/simpleeve/eo-skills/eo-change-review"><img src="https://agentmods.dev/badge/skills/simpleeve/eo-skills/eo-change-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00097 | $0.02111 |
| Opus 5 | $0.00048 | $0.01056 |
| Sonnet 5 | $0.00019 | $0.00422 |
| Haiku 4.5 | $0.00010 | $0.00211 |
Grade A, and why
eo-change-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
eo-change-review — Change 方案审查（按需闸门）
本 skill 是可选闸门：在 implement 前把牢「方向是否正确、AC 是否可验收、TODO 是否完整」。调用依据：../eo-shared/granularity.md §5 风险信号命中（且用户未豁免）或用户显式点名。
核心原则
- 方案级审查，不替作者做决定：只产出报告，修订由用户回
/eo-change执行 - AC 是重中之重：change 的价值密度集中在 §2；AC 不可验收，后面全白做
- 不审实施质量：此时代码还没写；维度 7 的前提校验落到真实代码（state 文档可能过期，读码取证不等于审码）
- 简版报告：
eo-doc/changes/<change-id>/change-review.md，复审覆盖重写（历史由 git 兜） - 只有 P0 阻塞：P1/P2 移交起草方裁决，不阻塞流程
对抗立场
默认这份方案通不过——审查的目标是找出让它失败的方式：构造会让 AC 失效的输入、会让 TODO 卡死的依赖、不成立的前提假设。「没发现问题」不是结论，「试图推翻它的方式都失败了」才是——报告须能列出试过的攻击面；列不出攻击面的审查视为没做。
前置条件
- 必须能找到
.eo-project.json。找不到 → 报错退出，提示运行/eo-project-init eo-doc/changes/<change-id>/change.md存在，status 为draft或confirmed（implementing 及之后 → 提示应走 /eo-review）
工作流程
第一步：阅读上下文
- 读目标 change.md 全文（§1 意图 + 已钉决策、§2 AC、§5 TODO、§6 风险）
- 读
eo-doc/changes/INDEX.md最近 3 条（避免与在途/已归档 change 冲突或重复） - 系统现状定位（校验变更前提）：
.codegraph/索引存在则codegraph explore优先召回；不存在则按目录收敛 + 源码直读 - 复审轮另加：读上一版 change-review.md 的 Finding 清单（核销后再找新 finding）
第二步：审查（7 维度）
- 维度 1 · AC 质量（最关键）：对照 ../eo-shared/ac-spec.md 逐条检查——用户视角？可独立验证？技术无关且可度量（无「正常工作」类主观词）？覆盖异常路径？refactor 类是否写了「行为不变」的回归口径？
- 维度 2 · TODO↔AC 映射：每条 TODO 标注了对应 AC 且映射成立？每条 AC 至少被一条 TODO 覆盖？越界 TODO 或悬空 AC → P0
- 维度 3 · TODO 拆解质量：三要素齐全？占位符检测（「补充错误处理」「后续完善」→ P0）？Batch 分组合理、Batch 1 是可独立验证的 MVP？并行组核验（存在字母后缀批时）：文件集相交或有逻辑依赖 → P1
- 维度 4 · 粒度合规：对照 granularity.md——超软标 P1 建议拆、超硬标 P0 必须拆；反向检查：trivial 到不该开 change → P1 建议转直改
- 维度 5 · 意图一致性：§1 已钉决策与 §2/§5 自洽？
type与实际内容匹配？AC/TODO 超出 §1 意图自行扩面（镀金）→ P1 建议裁剪或转 backlog；§2 是否真是用户口吻（复述实现细节 → P1） - 维度 6 · 条件节合规：命中风险信号却未在 §6 记录（含豁免）→ P1；defer 超 3 条 → P1
- 维度 7 · 前提真实性抽查：§1/§5/§6 中对现状事实、兼容性、唯一性、外部契约的断言，按不可逆性 > 数据/权限风险 > 跨模块面排序取 2-3 条定点读码核验。每条前提标基线（变更前事实审
base_commit或 HEAD）。前提不成立 → P0（必须附直接证据 file:line）；证据不足 → P1 要求补证，不猜 P0
定级纪律：
- P0 只收客观可判项：TODO↔AC 映射断裂 / 占位符 / 粒度超硬标 / AC 不可验证 / TODO 推翻已钉决策 / 前提不成立（附直接证据）。程度与取舍类判断最高 P1
- 不确定就降级：拿不准 P0 报 P1，拿不准 P1 报 P2；每条 P0/P1 必须落到具体位置并附可执行修复建议，给不出的不报
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago Changed · -154 lines · +18 tokens per session 2b43dcd30dd0
- 11d ago First seen · 265 lines · 79 tokens per session scan A d6080f95f103
eo-change-review is a skill published in the GitHub repository SimpleEve/eo-skills (31 stars, last pushed 4d ago), licensed MIT. It adds 97 tokens to every session and 2,111 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
refactor
Expert code refactoring based on Martin Fowler's catalog — improve maintainability without changing behavior. Covers code smells, composing methods, moving features, organizing data, simplifying conditionals, method calls, and generalization. Triggers on: refactor, 重构, clean up, improve code, code smell, extract…
understand
Review and understand freshly AI-generated code. Scans the current repo's uncommitted/branch changes and builds a Claude-style light-themed two-column webpage — left is the project folder tree of changed files, right is the selected file's syntax-highlighted diff (additions/deletions clearly distinguished from…
technical-debt-analyzer
Detect and analyze areas with high maintenance cost, poor design, or accumulated technical debt. Use this skill when reviewing codebases for quality issues, planning refactoring efforts, conducting code audits, assessing project health, identifying maintenance hotspots, or prioritizing technical improvements. Analyzes…
code-smell-detector
Identify and report code smells indicating poor design or maintainability issues in Python code, including duplicate code, magic numbers, hardcoded values, God classes, feature envy, inappropriate intimacy, data clumps, primitive obsession, and long parameter lists. Use when conducting code quality audits, preparing…
design-smell-detector
Identify design quality issues in code including high coupling, low cohesion, God classes, long methods, and other code smells. Use when: (1) Reviewing code architecture and design quality, (2) Identifying refactoring opportunities, (3) Detecting God classes or classes with too many responsibilities, (4) Finding high…
code-review-assistant
Conduct comprehensive code reviews identifying bugs, security issues, performance problems, code quality concerns, and best practice violations. Use when reviewing pull requests, examining code changes, evaluating new code, assessing code quality, or providing feedback on implementations. Analyzes code for…