om-auto-continue-pr
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external sources including Pull Request descriptions, review comments, and linked issues. The instructions in
references/agentic-setup.mdandreferences/adopt-pr.mdestablish a clear 'Untrusted content boundary', directing the agent to treat this information as data rather than instructions and to specifically flag directives like 'ignore previous instructions' as suspected prompt injection. - [COMMAND_EXECUTION]: The skill uses shell commands to manage isolated git worktrees and run project-specific validation gates. These operations are essential to the skill's function and include security measures such as validating and quoting externally-sourced values before shell interpolation.
- [CREDENTIALS_SAFE]: The skill implements strict hygiene rules for sensitive information, explicitly requiring the redaction of tokens, API keys, and environment variables from any logs, plans, or comments generated by the agent.
Audit Metadata