The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is configured to read and adopt "patterns", "conventions", and "learnings" from AGENTS.md files located within the user's project structure to guide its research and documentation activities.
Ingestion points: The agent ingests data from local AGENTS.md files across various directory depths using the Read tool.
Boundary markers: The instructions lack explicit delimiters or warnings to prevent the agent from following malicious instructions that might be embedded within the project's documentation files.
Capability inventory: The skill possesses the Read, Edit, and Glob tools, which allows it to discover and modify files based on the context it has ingested.
Sanitization: No validation or sanitization process is implemented to ensure that the content being read from the repository's files does not contain override commands or deceptive instructions.
[EXTERNAL_DOWNLOADS]: The skill makes use of the mcp__perplexity-ask__perplexity_ask tool to retrieve external information during the best practices research phase.
This tool facilitates research by querying Perplexity AI, an established and well-known search and discovery service.

research-compound