ddd-book
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill manages project-specific engineering documentation and manifests. All operations are confined to the local filesystem, specifically within the .ddd/ directory. No malicious behaviors such as data exfiltration, command injection, or unauthorized access were identified.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as ADRs, requirements documents, and project configuration files to populate the DDD manifest and knowledge map. This ingestion creates a surface for indirect prompt injection. However, given the skill's restricted capabilities—limited to generating documentation artifacts—the overall risk is considered safe.
- Ingestion points: Project manifest files (package.json, lockfiles), architectural decisions (ADRs), requirements, and project documentation (CONTEXT.md, AGENTS.md) as listed in SKILL.md.
- Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore' instructions for processed content.
- Capability inventory: Directory creation and writing of YAML configuration/manifest files within the .ddd/ path.
- Sanitization: No sanitization or content validation for the indexed documents is defined.
Audit Metadata