The Agent Skills Directory

[SAFE]: No malicious behavior or code was detected. The skill is composed of markdown instructions and evaluation data.
[PROMPT_INJECTION]: Indirect Prompt Injection Surface: The skill instructs the agent to ingest and analyze untrusted data from external sources such as Reddit, Quora, and competitor blogs. This is a common pattern for research skills but represents an entry point for external instructions.
Ingestion points: SKILL.md (Web searches on Reddit, Quora, and competitor sites; user transcripts).
Boundary markers: None defined.
Capability inventory: Agent web search and file reading.
Sanitization: None specified.

content-strategy