api-mocking
Warn
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill identifies, downloads, and executes code from remote Postman workspaces. It uses
postman dependency add mock <mockId>to fetch a mock andpostman mock runto execute it locally as a Node.js server. The instructions explicitly state that the mock is "code that runs as you." - [DYNAMIC_EXECUTION]: The skill employs
postman mock generate <spec or collection>to automatically produce Node.js handler scripts (e.g.,default.js) based on external API definitions. These scripts are subsequently executed locally, representing a dynamic code generation and execution flow. - [INDIRECT_PROMPT_INJECTION]: There is a potential for indirect injection through the processing of untrusted API specifications or collections.
- Ingestion points: External API specifications or Postman collections processed by
postman mock generate(SKILL.md). - Boundary markers: None present in the instructions to prevent the generated code from being influenced by malicious content within the specification.
- Capability inventory: The generated code runs as a local Node.js process with access to the user's environment and standard libraries.
- Sanitization: No sanitization or validation of the input specification is described before it is used to generate executable code.
- [COMMAND_EXECUTION]: The skill relies on the
postmanCLI to perform various sensitive operations including searching for workspaces, listing mocks, and deploying them to remote servers. - [EXTERNAL_DOWNLOADS]: The skill downloads mock implementations and dependencies from external Postman workspaces via the
postman dependency addcommand.
Audit Metadata