neckbeard
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is architected to process external data from issue trackers, ticket systems (such as Linear, Jira, and Notion), and email threads as its primary intake mechanism.
- Ingestion points: Untrusted data enters the agent context during Phase 1 (Intake and provenance) as defined in
references/journey.mdand Phase 1 (Snapshotting) as defined inreferences/lifecycle.md. - Boundary markers: The skill instructions do not explicitly mandate the use of strict delimiters (e.g., XML tags, triple quotes) or "ignore embedded instructions" warnings when the agent reads or snapshots external issue bodies and comments.
- Capability inventory: The skill environment, as described in
SKILL.md, grants the agent file read/write capabilities and terminal access to perform domain implementation, verification, and repository management. - Sanitization: There is no explicit requirement for the agent to sanitize, escape, or validate external content before it is interpolated into the context for planning or execution stages.
Audit Metadata