Back to skill
Skillv1.0.1

ClawScan security

App Store Changelog · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignFeb 11, 2026, 8:21 AM
Verdict
benign
Confidence
high
Model
gpt-5-mini
Summary
The skill's files and runtime instructions are consistent with its stated purpose: it collects git commits/files and helps draft App Store release notes, and it does not request credentials, downloads, or other unexpected capabilities.
Guidance
This skill appears coherent and limited to reading the repository's git history and helping write release notes. Before running or granting the agent permission to execute it: (1) ensure you run it from the intended repository root so it doesn't read another project; (2) review commit messages and the generated notes for any internal-only wording or accidentally committed secrets (commit messages can leak sensitive info); (3) run the included script locally first to inspect its output; and (4) always review and edit the final App Store text before publishing. If you allow autonomous agent execution, remember the agent will be able to read the repo's git history when invoked.

Review Dimensions

Purpose & Capability
okThe name/description say it will gather git history and produce App Store release notes. The included script uses only git commands to list commits and touched files; the SKILL.md describes triage and drafting steps that map directly to that data. There are no unrelated requirements (no cloud creds, no external services).
Instruction Scope
okSKILL.md instructs the agent to run the provided script from the repository root, review commit messages and touched files, group user-facing changes, and produce brief bullets. Those instructions stay within the stated purpose. Note: commit messages and diffs may contain internal details or secrets — the skill's guidance to drop internal-only work is appropriate but the user should still review generated notes for sensitive info.
Install Mechanism
okThere is no install spec; this is instruction-only plus a small bundled bash script. Nothing is downloaded or written to disk by an installer, and the script itself is short, readable, and uses only standard git tooling.
Credentials
okThe skill declares no required environment variables, credentials, or config paths and the runtime instructions do not attempt to read secrets or unrelated env vars. The script accesses only the local git repository, which is proportional to the task.
Persistence & Privilege
okThe skill does not request permanent presence (always:false) and does not modify other skills or system-wide settings. It simply reads git history and prints results; it does not persist new credentials or alter agent configuration.