deep-build
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external, untrusted data including project instructions, current system state, and build contracts.
- Ingestion points: Found in
SKILL.mdwhere the agent is instructed to read the target project's instructions and current state. - Boundary markers: The instructions lack specific delimiters or instructions to ignore potential commands embedded within the external documents.
- Capability inventory: The agent has the authority to modify code, execute verification tasks, and perform UI interactions as outlined in
SKILL.mdandagents/openai.yaml. - Sanitization: There are no explicit requirements for sanitizing or validating the input from these external sources before the agent acts on them.
Audit Metadata