Skillv1.0.0

ClawScan security

Github Actions Linter · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 13, 2026, 1:11 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill is internally consistent: it bundles a pure-Python linter script and instructions to run it against GitHub Actions workflow YAML files and does not request unrelated credentials, installs, or persistent privileges.
Guidance: This skill appears to be a straightforward, bundled Python linter for GitHub Actions workflows. Before installing or running it: (1) inspect the full script (scripts/gha_linter.py) yourself — running arbitrary code has risk even if it looks benign; (2) run it on a local or sandboxed copy of your repository if you have sensitive workflows or secrets (the tool will read any files you point it at); (3) expect some false positives/negatives because it uses a custom minimal YAML parser; (4) no credentials are required, so avoid granting tokens or secrets to the skill. If you want extra assurance, run the script in an isolated environment (container) and review its full source for any network or shell operations before use.

Purpose & Capability: okName/description (GitHub Actions linter) match the delivered artifacts: runtime instructions call the bundled Python script and the repository contains a linter implementation. There are no unrelated required binaries or credentials.
Instruction Scope: noteSKILL.md instructs the agent to run the included script (python3 scripts/gha_linter.py) against one or more files or directories and provides modes (lint, security, deprecated, validate). This necessarily reads workflow files (e.g., .github/workflows/*.yml) and can be pointed at arbitrary paths — expected for a linter, but users should be aware it will parse any files/dirs you ask it to analyze.
Install Mechanism: okNo install spec or external downloads. The linter is bundled as a local Python script and claims to rely only on stdlib; nothing is pulled from external URLs or package registries.
Credentials: okThe skill declares no required environment variables or credentials. SKILL.md and the visible code do not reference secret or external credentials. Requesting no tokens is proportional to a static linter.
Persistence & Privilege: okalways is false and the skill does not request elevated or persistent presence. There are no instructions to modify other skills or global agent configuration.