changelog-generator

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes several shell commands to perform its tasks within the local repository environment.
  • Interacts with the filesystem using ls, cat, and grep to manage package.json and CHANGELOG.md files.
  • Uses standard git operations including git fetch, git log, and git show to extract metadata and code diffs from the repository history.
  • Executes bun scripts/version.ts patch to automate version bumping. This command triggers a local script assumed to be part of the repository's build tools.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes data from git commit messages and code diffs.
  • Ingestion points: Commit messages and code diffs retrieved via git log and git show in SKILL.md and references/writing-entries.md.
  • Boundary markers: None; there are no specific delimiters used to separate the external commit data from the agent's internal reasoning instructions.
  • Capability inventory: The agent has the ability to execute git commands, file system commands, and the local versioning script via bun.
  • Sanitization: No sanitization or filtering logic is provided to handle potentially malicious instructions embedded in commit messages.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 12:33 AM