agile-epic
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from repository files and user arguments to generate artifacts.
- Ingestion points: Data is read from project files in the 'planning/' directory, business rules, and the '$ARGUMENTS' variable.
- Boundary markers: The skill lacks explicit delimiters to distinguish between instructional templates and ingested content.
- Capability inventory: The skill generates 'Verification' sections in story files, which typically include executable shell commands.
- Sanitization: No sanitization or validation of ingested content is performed before generating task lists or verification steps.
Audit Metadata