comet-safe-delivery

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions involve executing standard Git commands like git status, git diff, git commit, and git push to manage repository delivery. These actions are within the intended scope and require explicit user authorization.
  • [SAFE]: The skill references a local file maintainer-contract.md for maintainer guidelines. This is an internal project reference and does not involve downloading content from external or untrusted sources.
  • [PROMPT_INJECTION]: The skill interacts with external data via repository diffs and Git status outputs, creating a surface for indirect prompt injection. 1. Ingestion points: git diff, git status, and local reference files in SKILL.md. 2. Boundary markers: Explicit markers are absent, but the skill mandates user authorization for all state changes. 3. Capability inventory: The skill utilizes Git commands with file-write (commit) and network (push) capabilities. 4. Sanitization: No explicit sanitization of repository content is specified. The risk is managed through mandatory user oversight.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 03:00 PM
Security Audit — agent-trust-hub — comet-safe-delivery