The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill consists of documentation and usage guidelines for external tools.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external web content via Perplexity search and synthesis tools. This is the primary function of the skill and is managed by recommended constraints on token limits and result counts to minimize context bloat and potential injection surface.
Ingestion points: mcp__perplexity__perplexity_search and mcp__perplexity__perplexity_ask in SKILL.md and README.md.
Boundary markers: Not explicitly provided in the usage examples.
Capability inventory: Fetches and synthesizes information from external web sources.
Sanitization: Relies on the underlying AI model's safety filters for processed web content.

perplexity