correct
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and act upon potentially untrusted data from historical repository logs.\n
- Ingestion points: The agent is instructed to read "recent commits, reverts, review comments, agent instruction files, and comments that explain workarounds" in the file
SKILL.md.\n - Boundary markers: The instructions lack delimiters or warnings to ignore embedded directives within the ingested historical data, which could allow malicious instructions in commit messages to influence the agent's behavior.\n
- Capability inventory: The skill allows the agent to make significant changes to the repository, including modifying architecture, adding lints and CI checks, and updating test suites.\n
- Sanitization: No sanitization or validation mechanisms are specified for the ingested data, making the skill vulnerable to poisoning attacks where an attacker uses commit history to trigger unsafe refactoring actions.
Audit Metadata