write-openapi
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as its primary function involves processing external OpenAPI specifications which may contain malicious instructions.\n
- Ingestion points: OpenAPI YAML/JSON files provided via file paths or URLs as defined in the skill instructions.\n
- Boundary markers: Absent; the skill does not define specific markers or delimiters to isolate untrusted specification data from the agent's internal instructional context.\n
- Capability inventory: The agent can execute subprocesses via
curland thegitbookCLI, and it has network access to publish data to API endpoints.\n - Sanitization: Absent; there is no logic described to validate or sanitize the content of the specifications for adversarial prompts before processing.\n- [COMMAND_EXECUTION]: The skill provides instructions for using shell commands such as
curland thegitbookCLI. These are used to perform legitimate operations against the vendor's API endpoints (api.gitbook.com) and are essential for the skill's functionality.\n- [EXTERNAL_DOWNLOADS]: The skill references and utilizes the official GitBook MCP server and CLI tool. These resources are hosted on vendor-controlled infrastructure and are necessary for extending the agent's capabilities to interact with GitBook services.
Audit Metadata