The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests external API specifications to generate mock configurations, serving as an entry point for untrusted data that could potentially influence the agent's behavior.
Ingestion points: ContractMockServer.load_contract in resources/implementation-playbook.md reads from external contract files.
Boundary markers: No explicit boundary markers or instruction filtering are implemented in the contract parsing logic.
Capability inventory: The skill generates server logic, routes, and data fixtures based on external specifications.
Sanitization: No sanitization of the contract content is performed prior to processing.
[DYNAMIC_EXECUTION]: The skill includes templates for dynamic script generation, allowing the creation of runtime-modifiable behavior based on request data.
Evidence: The implementation playbook includes functions like create_dynamic_stub, create_pytest_integration, and implement_throttling that return executable Python code strings as templates.
[EXTERNAL_DOWNLOADS]: The documentation template references resources from a well-known CDN service for interactive Swagger UI rendering.
Evidence: The create_interactive_docs function in resources/implementation-playbook.md includes CDN links to unpkg.com for Swagger UI assets.

api-testing-observability-api-mock