The Agent Skills Directory

[SAFE]: The skill implementation is focused on legitimate data discovery and sampling tasks. It correctly identifies the risk of handling secrets and provides explicit guardrails to prevent the agent from leaking WordPress application passwords or WooCommerce keys into the command output. It follows the wix vendor standard for migration projects by organizing files within project-local directories.\n- [INDIRECT_PROMPT_INJECTION]: The skill functions as a data ingestor, creating a potential surface for indirect prompt injection by retrieving untrusted content from the WordPress REST API.\n
Ingestion points: External data is fetched from the target WordPress/WooCommerce site via the REST API using fetchJson in scripts/wp-discovery.js.\n
Boundary markers: Sample data is structured within markdown JSON code blocks in the discovery files generated by the writeOutputs function in scripts/wp-discovery.js.\n
Capability inventory: The tool utilizes fetch for network read operations (defined in lib/wp-http.js) and fs.writeFile for writing discovery artifacts (in scripts/wp-discovery.js).\n
Sanitization: The skill performs discovery by sampling raw records without modification, which is the intended behavior for ensuring data faithfulness during migration analysis.

rp-source-wordpress