make-ticket
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a standard markdown format (Problem, Outcome, Affected Users/Systems, Constraints, Open Questions) to help agents convert discussions or ideas into actionable tickets. The instructions emphasize clarity and intent over specific technical solutions.
- [INDIRECT_PROMPT_INJECTION]: The skill acts on untrusted data from user-provided discussion threads and repository source code during the 'grounding' phase. The risk of the agent following instructions embedded in these sources is low due to the strict five-section format requirements and the explicit guardrail stating the agent must never file a ticket without a user's go-ahead.
- [DATA_EXPOSURE]: The agent is instructed to search for authentication models and data-handling rules within the repository to identify ticket constraints. While this involves processing sensitive architectural information, the skill is designed to summarize these into high-level constraints rather than exfiltrating raw credentials or secrets, and the final output is subject to user review.
Audit Metadata