The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest data from local network endpoints and database schemas to inform API design. This presents a theoretical surface for indirect injection if these sources were compromised.
Ingestion points: Found in SKILL.md referencing http://localhost:8080/api-docs and database schema queries.
Boundary markers: None explicit for external content.
Capability inventory: SQL execution via Supabase MCP and file searching via Grep.
Sanitization: No specific sanitization mentioned for ingested documentation text.
[COMMAND_EXECUTION]: The skill utilizes SQL commands to inspect the database environment's structure.
Evidence: Templates for querying information_schema.columns and information_schema.table_constraints provided in SKILL.md to assist in design reasoning.
[EXTERNAL_DOWNLOADS]: The skill attempts to access local network resources for configuration and documentation.
Evidence: References to http://localhost:8080/api-docs and http://localhost:8080/naming-conventions for pre-design checks.

design-api