rest-api-builder
Warn
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The author field is set to 'Claude Code' (the platform name) instead of the actual developer 'inbharatai'. This deceptive labeling can lead users to believe the skill is a verified, official component of the agent platform when it is actually third-party content.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided API specifications to generate executable code, which creates a surface for malicious instructions to influence the agent's output.
- Ingestion points: User input defining API structure and requirements provided during runtime interaction.
- Boundary markers: No structural delimiters or specific instruction-isolation techniques are defined to separate user data from agent instructions.
- Capability inventory: The skill is capable of generating production-ready code for web frameworks, including authentication and rate-limiting logic.
- Sanitization: The instructions include a guideline to 'Always validate inputs before processing', which is a recommended defensive practice, though it focuses on the generated code rather than protecting the agent context itself.
Audit Metadata