The Agent Skills Directory

[COMMAND_EXECUTION]: The signal-scorer, mutual-mapper, and enrichment-agent are all configured with the Bash tool. This provides a broad capability for arbitrary command execution on the host environment, which is used here for data filtering (Grep, Glob) and general processing.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted data from external sources.
Ingestion points: Untrusted content enters the agent context via WebSearch, WebFetch, and X API results in enrichment-agent.md, signal-scorer.md, and mutual-mapper.md.
Boundary markers: The instructions do not define delimiters or warnings to ignore instructions embedded in the external data (e.g., in bios or tweets).
Capability inventory: The agents have access to Bash, Read, and WebFetch.
Sanitization: No sanitization or escaping of the external content is performed before it is used to draft outreach messages in outreach-drafter.md.
[EXTERNAL_DOWNLOADS]: The skill performs extensive network operations to fetch data from well-known services including the X API, Exa Search, and LinkedIn. While these are legitimate for the skill's purpose, they involve retrieving content from across the open web.

lead-intelligence