api-endpoint-builder
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to generate code that processes external, potentially untrusted user input via HTTP requests. This creates an inherent attack surface for indirect prompt injection if the generated endpoints are used to feed data back into an LLM or agent context without proper sanitization.
- Ingestion points: Request bodies, parameters, and query strings in generated route handlers as defined in
SKILL.md. - Boundary markers: The skill includes explicit instructions and templates to validate all fields before processing and suggests the use of schema validation.
- Capability inventory: The skill itself is purely instructional and does not possess active capabilities such as file system writing or network operations.
- Sanitization: The templates provided include detailed input validation logic, password hashing with bcrypt, and instructions for using parameterized queries and rate limiting to prevent common injection attacks.
Audit Metadata