The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local OpenAPI/Swagger files and remote Postman collections to drive code generation and security audits, creating a vulnerability to malicious instructions embedded in those data sources.
Ingestion points: The skill searches for local files matching **/openapi.{json,yaml,yml} and **/swagger.{json,yaml,yml} and fetches remote collection data using the getCollection and getSpecDefinition tools.
Boundary markers: The instructions lack explicit delimitation or "ignore embedded instructions" warnings when processing these external specifications.
Capability inventory: The skill utilizes Write permissions to generate client code files and has access to the Bash tool for command execution.
Sanitization: No sanitization or validation of the specification content is defined before it is used to influence the agent's output.
[DATA_EXPOSURE]: The "API Security Audit" workflow is designed to programmatically inspect Postman environment variables via the getEnvironment tool to detect leaked secrets. While intended as a security feature, it provides the agent with access to sensitive project credentials.
[PERSISTENCE]: The setup documentation instructs users to modify shell configuration files (~/.zshrc or ~/.bashrc) to persist the POSTMAN_API_KEY. While this is a common manual setup step for developers, it involves modification of sensitive system-level persistence files.

postman