ClawHub

code-modification-guard

v1.0.2

Code modification guideline / 代码修改规范指南。Only make changes explicitly authorized by the user / 仅执行用户明确授权的改动。Never simplify, delete, or rename existing code ele...

⭐ 0· 38·0 current·0 all-time
by小熙@dreamersetsail
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description match the instructions: the skill's purpose is to constrain code modifications and the SKILL.md contains rules and checks that directly implement that purpose. It does not request unrelated credentials, binaries, or installs.
✓
Instruction Scope
The instructions tell the agent to analyze project files and search for reusable resources before making changes — this is appropriate and expected for a code-modification guard. The doc does not instruct access to unrelated system paths, secrets, or external endpoints.
✓
Install Mechanism
No install spec or code files are present; this is instruction-only so nothing is written to disk or fetched during install.
✓
Credentials
The skill requires no environment variables, credentials, or config paths. The behaviors described (reading project structure, asking for authorization) are proportional to its stated function.
✓
Persistence & Privilege
always is false, autonomous invocation is allowed (platform default). The skill does not request permanent presence, nor does it modify other skills or system-wide settings.
Assessment
This skill is coherent and restrictive by design: it instructs the agent to analyze project files and to only apply changes after explicit user authorization. Before installing, confirm you want the agent to follow these rules automatically whenever code changes are discussed (the skill requests that behavior). Also consider operational safeguards: require the agent to show diffs and ask for explicit approval before writing files, avoid granting broad remote/network access to the agent, and test the agent's behavior on a small repo to ensure it asks for confirmation as promised.

Like a lobster shell, security has layers — review code before you run it.

latestvk978sn4vbjwfqcg3e7sz4zgw7h844s91

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments