The Agent Skills Directory

[SAFE]: The skill communicates with its own vendor infrastructure (api.aisa.one) to perform search and synthesis operations.
[SAFE]: Credentials are managed securely via environment variables (AISA_API_KEY) as per standard practices for agent skills.
[SAFE]: No suspicious external dependencies, remote code execution, or obfuscation techniques were detected. The script uses only Python standard libraries.
[INDIRECT_PROMPT_INJECTION]: Like all tools that ingest data from external web sources, the skill has a surface for indirect prompt injection where instructions could be embedded in search results. However, the skill treats this data as text for display or synthesis and does not use it to drive high-privilege operations.
Ingestion points: Web search results, academic paper snippets, and Tavily search content.
Boundary markers: Output is formatted into clear text blocks with source labels.
Capability inventory: Limited to network GET/POST to vendor API and stdout printing.
Sanitization: Result snippets are truncated before display to prevent excessive context injection.

multi-search