The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is designed to ingest and process data from untrusted external sources, such as business websites and third-party directories, through tools like Firecrawl and Browserbase.
Ingestion points: Data is programmatically extracted from prospect websites (e.g., About pages, team pages) and public listings to qualify leads.
Boundary markers: The instructions do not specify the use of clear delimiters or instructions for the agent to disregard potentially malicious commands embedded in the scraped web content.
Capability inventory: The skill has the capability to generate files (CSV lead sheets) and perform network operations via multiple integrated prospecting APIs.
Sanitization: There is no documentation of specific sanitization or filtering logic applied to external data before it is incorporated into the agent's context.
[COMMAND_EXECUTION]: The skill references and provides instructions for executing a local Node.js utility script, github-prospects.js, located within the tools/clis/ directory. This script is used to collect and enrich data from GitHub repositories based on user-supplied parameters.
[EXTERNAL_DOWNLOADS]: The skill documentation refers to several external data providers and SaaS platforms for firmographic enrichment, contact discovery, and email validation (e.g., Apollo, ZoomInfo, Clay, Truelist, Hunter). These are well-known industry services used for legitimate business prospecting workflows.

prospecting