The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or security violations were detected. The skill is designed for legitimate company research tasks.
[DATA_EXFILTRATION]: The skill utilizes official Exa API tools (e.g., agent_create_run, web_search_advanced_exa) to retrieve web data. All external links point to the vendor's official documentation at docs.exa.ai, and there is no evidence of unauthorized data harvesting.
[PROMPT_INJECTION]: While the skill processes content from the external web, it explicitly instructs the agent to maintain 'Token Isolation.' This practice prevents untrusted raw data from being dumped directly into the primary context, mitigating potential indirect prompt injection attacks.

company-research