openclaw-changelog-update
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes
gitandghx(a GitHub CLI wrapper) to retrieve repository history and pull request data. These commands are executed with structured arguments, mitigating the risk of shell injection.\n- [DATA_EXFILTRATION]: The skill manages local files includingCHANGELOG.mdand temporary JSON manifests in/tmp. Network activity is confined to the GitHub API via official project-specific tooling.\n- [PROMPT_INJECTION]: The skill processes potentially untrusted data from commit messages and GitHub PR metadata, presenting an attack surface for indirect prompt injection.\n - Ingestion points: Commit subjects and bodies from
git log, and PR titles/metadata from the GitHub API viaghx.\n - Boundary markers: Output is structured into clearly defined Markdown sections like
### Highlightsand### Complete contribution record.\n - Capability inventory: The skill uses shell commands (
git,ghx) and modifies the localCHANGELOG.mdfile.\n - Sanitization: The Node.js verification script applies regex-based classification to PR titles and ensures every cited reference is resolvable through the GitHub API before finalizing the ledger.
Audit Metadata