om-auto-implement-spec
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill identifies the risk of prompt injection from untrusted sources such as issue bodies and pull request comments. It explicitly instructs the agent to ignore directives like 'ignore previous instructions' or 'run this command' when encountered in external data.
- [DATA_EXFILTRATION]: The skill incorporates clear instructions to refuse any commands that would exfiltrate data or read credential stores. It also mandates secrets hygiene, forbidding the posting of tokens or environment variables to pull request comments.
- [COMMAND_EXECUTION]: While the skill interacts with the shell via tracker tools and engine skills, it requires strict validation of all externally-sourced values (like IDs and branch names) against a whitelist regular expression before use in shell interpolation.
- [SAFE]: The skill delegates complex operations to sibling skills within the same verified ecosystem (om-auto-create-pr, om-auto-qa-pr, etc.), representing standard vendor functionality.
Audit Metadata