The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process external data such as OpenAPI specifications, Postman collections, and curl commands without explicit boundary markers.
Ingestion points: According to prompts/api-test-postman.md, the skill accepts OpenAPI files, curl commands, and Postman collections as input.
Boundary markers: There are no delimiters or instructions provided to the agent to treat external content as untrusted data.
Capability inventory: The skill produces executable Postman collections and includes a helper script (scripts/run-tests.sh) for running the newman CLI tool.
Sanitization: No evidence of input sanitization or validation was found in the instructions.
[COMMAND_EXECUTION]: The scripts/run-tests.sh script executes the newman command-line tool. It utilizes shell variables ($COLLECTION and $ENVIRONMENT) to define execution paths, which could be exploited for command injection if the environment is not strictly controlled.
[PRIVILEGE_ESCALATION]: The README.md file advises users to use the -ExecutionPolicy Bypass parameter when installing the skill via PowerShell. This practice encourages users to circumvent local security configurations that restrict script execution.

api-test-postman