konnect-api-catalog
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data in the form of API specifications, documentation, and catalog metadata through the
kong-konnectMCP server and user-provided artifacts. - Ingestion points: API specifications, documentation pages, and metadata fields are processed across all workflow steps described in
SKILL.mdand the reference files. - Boundary markers: The instructions do not define specific delimiters or "ignore" instructions for content extracted from these external files.
- Capability inventory: The skill identifies capabilities to interact with the environment via
kong-konnect(MCP),terraform-konnect,kongctl-declarative, anddeck-gatewayto read and potentially modify Konnect configurations. - Sanitization: There are no explicit instructions to sanitize or validate the content of descriptions or metadata fields before including them in the diagnostic analysis.
Audit Metadata