spec-dd
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automatically detects and executes shell commands to run tests and verify code compliance.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes project source code and existing documentation, creating a surface for indirect instructions.
- [INGESTION_POINTS]: Reads implementation code and specifications as described in SKILL.md (Phase 6 and 7).
- [BOUNDARY_MARKERS]: Absent; there are no specific markers to distinguish between data and instructions during code review.
- [CAPABILITY_INVENTORY]: Subprocess execution via shell commands for test running as documented in references/review.md.
- [SANITIZATION]: Absent; the skill analyzes file content directly without filtering for potential prompt injection.
Audit Metadata