openapi-format
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill enables processing of untrusted OpenAPI specifications from local or remote sources. These documents may contain malicious instructions in text fields (e.g., descriptions) intended to hijack the agent's logic.\n
- Ingestion points: The
oaFileargument allows reading from files or URLs as stated inSKILL.md.\n - Boundary markers: No delimiters or ignore-instructions warnings are specified for the agent.\n
- Capability inventory: The
openapi-formatCLI can read/write local files and perform network requests.\n - Sanitization: No input validation or content filtering is implemented for the processed documents.\n- Data Exposure & Exfiltration (LOW): The
--playgroundoption sends the content of the document to an external third-party service for link generation. Evidence found inreferences/command-recipes.md.\n- External Downloads (LOW): TheoaFileparameter supports fetching specifications from arbitrary remote URLs, potentially downloading untrusted content. Evidence found inreferences/feature-matrix.md.\n- Command Execution (SAFE): The skill is designed to run theopenapi-formatCLI tool for API document transformation, which is the primary and documented intent of the skill.
Audit Metadata