to-issues
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill establishes a structured, read-to-write workflow for task management. It explicitly requires a user-sign-off phase ('Quiz the user') before the agent performs any write operations to an external issue tracker, ensuring human oversight.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes data from external project documents and issue trackers.
- Ingestion points: External issue bodies, comments, and project files are read during context gathering and codebase exploration in
SKILL.md. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the processing logic.
- Capability inventory: The agent can publish content to an external issue tracker (Step 5 in
SKILL.md). - Sanitization: The skill does not define specific sanitization or filtering for the external content it processes.
- Mitigation: The risk is minimized by the interactive feedback loop in Step 4, where the user must approve the drafted issues before publication.
Audit Metadata