doubaoya-gateway
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill facilitates network requests to the vendor's domain (doubaoya.com) for authenticated API operations.
- Security measures are explicitly mandated: API keys (DOUBAOYA_API_KEY) must be retrieved from environment variables and must not be logged, echoed, or displayed in full.
- The network destinations are recognized vendor resources associated with the skill's primary function as a gateway (File: SKILL.md).
- [INDIRECT_PROMPT_INJECTION]: The skill design relies on dynamically fetching API contracts and UI schemas from the remote server at runtime, creating an ingestion surface for external instructions.
- Ingestion points: Data enters the agent's context through GET requests to https://doubaoya.com/api/apis/ and https://doubaoya.com/api/skills/ (File: SKILL.md §1).
- Boundary markers: The protocol utilizes a structured response envelope ({ success, requestId, data, error }) and recommends validating all inputs against a server-provided json-schema.
- Capability inventory: The skill enables the agent to perform GET, POST, and PUT operations based on target paths returned by the API server (File: SKILL.md §3.3).
- Sanitization: The skill mandates following dynamic json-schema (zod-derived) and inputUiSchema constraints, which define specific fields, types, and constraints to prevent unexpected parameter values.
Audit Metadata