api-portal-discovery
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted external data from OpenAPI specifications and GraphQL schemas which could be used for indirect prompt injection.
- Ingestion points: WebFetch tool responses from discovered subdomains and paths.
- Boundary markers: Absent.
- Capability inventory: Bash and WebFetch tool execution.
- Sanitization: The skill contains a directive to not execute arbitrary code from API specs, reducing the risk of accidental execution of instructions embedded in data.
- [COMMAND_EXECUTION]: The skill uses Bash for discovery operations and executes local hook scripts for tool lifecycle management.
Audit Metadata