The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill demonstrates processing untrusted data from the web (search results from Exa.ai), which could potentially contain malicious instructions.
Ingestion points: exa.search_and_contents results, specifically the text, summary, and highlights fields retrieved from the web in company-search.md, people-search.md, and websets.md.
Boundary markers: None are present in the provided code snippets to distinguish between instructions and data.
Capability inventory: The skill contains code for local file system access, specifically writing search results to CSV and other files using open() in company-search.md and websets.md.
Sanitization: The examples do not include sanitization or validation of the retrieved web content before it is printed or written to disk.
[Data Exposure & Exfiltration] (LOW): The skill includes code snippets that write data to the local file system (e.g., export_companies_to_csv in company-search.md). While appropriate for the skill's stated purpose of lead generation, this is a capability that could be misused if modified.

exa-entities