The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references external resources for tool installation, including the belt CLI and an installation guide hosted on raw.githubusercontent.com/inference-sh.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting untrusted data from external search services (Tavily and Exa) via the belt CLI. The retrieved data is then processed and interpolated into the persona templates, which presents a surface for potential indirect prompt injection if the search results contain malicious instructions.
Ingestion points: External search results from tavily/search-assistant, exa/search, and exa/answer are piped into the agent's persona construction workflow.
Boundary markers: None explicitly defined in the provided instruction patterns.
Capability inventory: Uses the Bash tool to execute belt commands which interact with the network and external APIs.
Sanitization: No visible sanitization or filtering of the external search content before it is processed by the agent.

customer-persona