resolve-todo-parallel
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The references/evals.yaml file contains simulated injection attempts within test cases to verify the agent's refusal behavior. These are data points for testing and do not affect runtime instructions.
- [DATA_EXFILTRATION]: The skill mandates the redaction of secrets and credentials from all outputs, preventing accidental exposure of sensitive information during the todo resolution process.
- [COMMAND_EXECUTION]: Destructive operations like file deletion and git mutations (commit/push) are strictly gated by mandatory user approval, preventing unauthorized or accidental changes to the repository.
- [PROMPT_INJECTION]: The skill manages indirect injection risks from the todos/ directory (SKILL.md) by requiring prior analysis of todo metadata and enforcing secret redaction. Its capabilities include file modifications and repository updates, which are controlled by explicit validation gates.
- [SAFE]: The skill implements a protected artifact policy that prevents the automated cleanup of important project documentation directories like docs/brainstorms/, docs/plans/, and docs/solutions/ unless an explicit user override is provided.
Audit Metadata