Back to skill
Skillv0.3.0

ClawScan security

Git Workflow · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignApr 27, 2026, 1:33 AM
Verdict
benign
Confidence
high
Model
gpt-5-mini
Summary
The skill's code, instructions, and requirements align with a Git workflow helper; nothing requests unrelated credentials or installs arbitrary code.
Guidance
This skill appears to be what it says: a Git/PR/release workflow helper with a small local validator script. Before using: 1) Be aware the skill's default is to push commits immediately without asking — if you need confirmation, change the workflow or ask the agent to confirm before pushing. 2) For PRs/releases you must authenticate gh (gh auth login) — the skill does not request extra credentials. 3) Review commits locally before invoking the skill in sensitive repositories (secrets or private branches) because it can perform remote pushes and create releases. Otherwise the files and instructions are proportionate and contain no hidden endpoints or unexpected installs.

Review Dimensions

Purpose & Capability
okName/description match the requested binaries and included validator script. Requiring git (and optionally gh) is appropriate for commits, PRs, and releases. No unrelated tools or credentials are requested.
Instruction Scope
noteInstructions stay within git/gh operations and use a local validator script. One noteworthy policy: the skill's default behavior is to 'always push immediately' after committing and 'Do not ask.' That is coherent with the stated workflow but may be surprising or risky for users who expect confirmation before network pushes—it gives the agent permission to perform remote writes without an extra confirmation step.
Install Mechanism
okNo install spec; instruction-only with a small included Python validator. Nothing is downloaded or written to disk by an installer step.
Credentials
okNo environment variables or credentials declared. The optional use of gh is noted and would require gh auth login, which is reasonable for PR/release actions. No unrelated secrets or config paths are requested.
Persistence & Privilege
okalways is false and the skill does not request persistent/platform-wide privileges. It does allow autonomous invocation (platform default), which combined with the 'auto-push' behavior is the primary operational risk to be aware of.