context-driven-development
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow centered on the agent reading and maintaining project artifacts like
product.mdandtech-stack.md. This creates a surface where instructions embedded in these files by other contributors or external sources could influence the agent's actions.\n - Ingestion points: Files located within the
conductor/directory and existing project source code analyzed during setup.\n - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are established for the agent when processing these artifacts.\n
- Capability inventory: The skill facilitates file system read/write operations for documentation management and utilizes the
/conductor:setupcommand functionality.\n - Sanitization: The skill lacks instructions for the agent to validate or sanitize the contents of the documentation files before incorporating the information into its operational context.
Audit Metadata