write-ticket
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from existing tickets and user prompts which could contain malicious instructions designed to override agent behavior during ticket generation.
- Ingestion points: Processes existing tickets and user-provided prompts during the investigation phase as described in SKILL.md.
- Boundary markers: While it uses a structured plan format (reference.md#plan), it lacks explicit markers or warnings to the agent to ignore instructions embedded within the data it analyzes.
- Capability inventory: The skill utilizes /analyze for file reading, /grill-me for user interaction, and the gh tool or Linear API for writing data to external trackers.
- Sanitization: No specific sanitization or filtering of input data is mentioned before it is incorporated into the final ticket description, though the 'grill-me' process requires the agent to restate intent in its own words, which acts as a partial defensive measure.
Audit Metadata