research-to-deploy
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill consumes untrusted data from the internet which could potentially influence agent behavior or configuration generation.
- Ingestion points: The skill instructions specify using WebSearch and WebFetch tools to retrieve documentation, community best practices, and known pitfalls from the public web.
- Boundary markers: The skill design includes a mandatory review phase where the agent must present a Research Summary (architecture rationale, cost estimates, security considerations) for user confirmation before generating any code artifacts.
- Capability inventory: The skill possesses the capability to Write and Edit files and execute Bash commands restricted to terraform, docker, kubectl, git, and npm operations.
- Sanitization: The skill relies on the human-in-the-loop review process as the primary mitigation for untrusted input, with no additional automated sanitization of web content mentioned.
Audit Metadata