The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and parses untrusted data from user-generated content like Linear comments and story descriptions. 1. Ingestion points: External comments with headers like '## Manual Testing Results' and '## Test Research'. 2. Boundary markers: Phase 3 uses regex-based structured parsing to delineate sections, providing basic logical boundaries. 3. Capability inventory: File system operations (read/write in '.hex-skills/') and execution of managed child worker skills (ln-301/ln-302). 4. Sanitization: No explicit sanitization of the parsed content is described before its use in generation. This surface is inherent to the functional purpose of a test planning tool and is managed via structured workflows.

ln-523-auto-test-planner