The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface by processing untrusted data from external sources such as scraped pages, reviews, and emails. It mitigates this risk with a specific instruction to disregard any embedded commands within that evidence. Ingestion points: SKILL.md Data Sources section. Boundary markers: Explicit instructions to treat scraped material as untrusted input. Capability inventory: No executable code or tools are defined within the skill. Sanitization: Instruction-based filtering.\n- [SAFE]: The skill consists entirely of markdown instructions without accompanying executable scripts or binaries. It adheres to security best practices by instructing the agent to keep API keys out of source control and to obtain explicit user authorization for any persistent or external actions.

performance-analyzer