api-test-pytest
Warn
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The script
scripts/generate_pytest_tests.pyis vulnerable to code injection. It takes API paths from external definitions (e.g., OpenAPI, Swagger, curl) and interpolates them directly into Python f-strings to generate test code. An attacker-controlled API definition with a crafted path containing quotes and comments (e.g.,/v1/path', ... #) could execute arbitrary Python code during the subsequent test execution. - [COMMAND_EXECUTION]: The
scripts/run.shscript automates a pipeline that generates Python code and immediately executes it usingpytest. This execution flow is dangerous when combined with the unsanitized input processing in the generator. - [COMMAND_EXECUTION]: The skill documentation refers to external installation scripts (
install-skills-mac.shandinstall-skills-windows.ps1) to be executed from the repository root. These scripts are not part of the provided skill files, representing an unverified execution of shell commands during setup. - [PROMPT_INJECTION]: The skill processes untrusted input files (API docs, curl commands) as part of its core functionality. While the logic in
prompts/api-test-pytest.mdis standard, the ingestion of external data without strict boundary markers creates a surface for indirect prompt injection.
Audit Metadata