The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an OpenAPI specification file fetched from a URL. * Ingestion points: scripts/openapi_analyzer.py loads the JSON specification from /tmp/mikopbx_openapi.json. * Boundary markers: No explicit delimiters are used to protect the agent from instructions embedded in spec fields like summary or description. * Capability inventory: The skill uses Bash, Read, Grep, and Glob tools. * Sanitization: The processing script does not sanitize the contents of the specification strings.
[DYNAMIC_EXECUTION]: An integration example in examples/integration-examples.md demonstrates generating Python test scripts by interpolating fields from the OpenAPI specification directly into a code template. This creates a risk of code injection if the analyzed specification contains malicious strings.
[COMMAND_EXECUTION]: The skill uses docker exec and curl to fetch data from the environment. This is intended functionality for the vendor's ecosystem.
[EXTERNAL_DOWNLOADS]: Fetches an OpenAPI specification from vendor-specific URLs (e.g., http://mikopbx-php83.localhost:8081/pbxcore/api/v3/openapi:getSpecification).

openapi-analyzer