← 返回 Skills 市场
Review Ai Writing
作者
Kevin Anderson
· GitHub ↗
· v1.0.2
· MIT-0
146
总下载
0
收藏
1
当前安装
3
版本数
在 OpenClaw 中安装
/install review-ai-writing
功能描述
Detect AI-generated writing patterns in developer text — docs, docstrings, commit messages, PR descriptions, and code comments. Use when reviewing any text a...
安全使用建议
This skill appears coherent and aligned with its stated purpose. Before installing, confirm the environment has git and (if you want PR scanning) the GitHub CLI (`gh`) available — the SKILL.md uses those but the skill doesn't declare them. Be aware the skill will read repository files, commit history, and optionally PR bodies, and will write a JSON report under .beagle/ai-writing-review.json (and will read .beagle/llm-artifacts-review.json if present). Also review/trust any referenced companion skills (beagle-core:review-verification-protocol and beagle-docs:review-ai-writing) if they exist in your environment. If you want to limit scope, run with the --category flag or avoid --all; because disable-model-invocation is true, the skill must be invoked explicitly by a user action.
能力标签
能力评估
Purpose & Capability
The name/description claim (detect AI-generated patterns in docs, comments, commits, PRs) matches the SKILL.md and the reference documents. The runtime steps (find relevant files, extract comments/docstrings, read commits and PR bodies, scan categories, and write a JSON report) are appropriate for this purpose. One minor mismatch: the instructions call `gh pr view --json body` and run git commands even though the skill declares no required binaries; the skill does not declare `gh` or `git` as dependencies.
Instruction Scope
Instructions explicitly read repository files, commit logs, and (optionally) PR bodies, then write results to a local .beagle/ai-writing-review.json. Reading/writing local repo files and running git/gh is consistent with the purpose. The skill also looks for an existing .beagle/llm-artifacts-review.json to avoid duplicate flags — this is sensible but means the skill will read that local file. No instructions attempt to read unrelated system files or environment variables, nor to exfiltrate data to external endpoints beyond calling `gh` (GitHub CLI) which is expected for PR body access.
Install Mechanism
This is an instruction-only skill with no install spec and no code files to install — lowest-risk install mechanism. There is no download or archive extraction. No concerns here.
Credentials
The skill requests no environment variables, credentials, or config paths. It only uses local repo state and (optionally) the GitHub CLI to read PR metadata. The lack of required credentials matches the intended scanning functionality.
Persistence & Privilege
Skill metadata sets disable-model-invocation: true (model cannot autonomously invoke it) and always: false. The skill writes a report to a repository-local directory (.beagle/) which is expected for a scanning/reporting tool. It does not request permanent platform-wide privileges or modify other skills' configurations.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install review-ai-writing - 安装完成后，直接呼叫该 Skill 的名称或使用
/review-ai-writing触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.2
- Documentation expanded for step 10 (Verification): final instruction now begins with "Before completing, all of the…" (was previously incomplete).
- No functional or behavioral changes; only documentation in SKILL.md file was updated.
- All other instructions and format remain unchanged.
v1.0.1
- Updated an autoContext keyword: replaced "humanize-ai-writing" with "humanize-beagle" for better alignment with related skill naming.
- Updated documentation in SKILL.md to reference the new `/beagle-docs:humanize-beagle` command and instructions.
- No changes to core functionality or command usage; only wording updates in the documentation.
v1.0.0
- Initial release of the AI Writing Review skill.
- Scans developer text—docs, code comments, docstrings, commit messages, and PR descriptions—for common AI-generated writing patterns.
- Detects issues across six categories: content, vocabulary, formatting, communication, filler, and code documentation.
- Automatically splits review into parallel agents to check prose, code comments/docstrings, and Git artifacts.
- Avoids duplicate findings by consolidating with existing llm-artifacts reviews.
- Generates a structured JSON report and a clear, actionable summary, including rewrite suggestions and fix safety guidance.
元数据
常见问题
Review Ai Writing 是什么？
Detect AI-generated writing patterns in developer text — docs, docstrings, commit messages, PR descriptions, and code comments. Use when reviewing any text a... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 146 次。
如何安装 Review Ai Writing？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install review-ai-writing」即可一键安装，无需额外配置。
Review Ai Writing 是免费的吗？
是的，Review Ai Writing 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Review Ai Writing 支持哪些平台？
Review Ai Writing 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Review Ai Writing？
由 Kevin Anderson（@anderskev）开发并维护，当前版本 v1.0.2。
推荐 Skills