The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill directs the agent to follow logic and sequences defined in external documentation files (blueprint/20-entrevista.md and blueprint/31-proveedores.md). This architecture allows instructions from secondary files to influence the agent's behavior during sensitive operations like credential rotation.
Ingestion points: blueprint/20-entrevista.md (Tramo 3), blueprint/31-proveedores.md (Meta, Zernio, Demo sections).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands in the referenced files are provided.
Capability inventory: State persistence to .wca-estado.json and guided modification of .env files.
Sanitization: No validation or sanitization of the content from the blueprint files is described.
[CREDENTIALS_UNSAFE]: The skill's primary purpose is the management of API credentials for various providers. It mitigates exposure risk by explicitly instructing the agent to never print secrets or copy them to other files, and mandates the use of .env files for secure storage.

conectar