Swagger Codegen Orchestrator
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEMETADATA_POISONINGINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [METADATA_POISONING]: The skill's frontmatter contains a
verification: security_reviewedfield. This is a self-authoritative claim intended to influence the perceived safety of the skill and should be disregarded in favor of independent analysis. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process OpenAPI 3.x specifications from multiple untrusted sources, including local files, arbitrary URLs, and the Swagger Hub API. This creates an attack surface where malicious instructions could be embedded within the documentation or schemas of the processed APIs to influence the agent's logic during the code generation process.
- Ingestion points: OpenAPI specs (local, URL, Swagger Hub API) referenced in
SKILL.md. - Boundary markers: None specified for handling external data.
- Capability inventory: Wraps
swagger-codegen-cliandopenapi-generatorto create files. - Sanitization: No explicit sanitization of input specifications is mentioned.
- [EXTERNAL_DOWNLOADS]: The installation instructions utilize
npx skills add agentskillexchange/skills. While this is the standard installation mechanism for the author's ecosystem, it involves downloading code from an external repository.
Audit Metadata