The Agent Skills Directory

[SAFE]: The skill is a complex but well-structured research orchestrator. It incorporates multiple layers of verification, including specialized fact-checker agents and cross-model reviews, and includes privacy-sensitive logic that adapts tool selection based on the sensitivity of the research topic.- [COMMAND_EXECUTION]: The skill utilizes the Bash tool for administrative tasks such as determining the project root directory and generating timestamps for workspace management. These operations are standard for maintaining organized local research artifacts.- [EXTERNAL_DOWNLOADS]: The skill fetches content from the web to fulfill research requests. It utilizes dedicated search and fetch tools as well as standard utilities for high-fidelity extraction, which is consistent with its primary research function.- [PROMPT_INJECTION]: As the skill processes untrusted content from the web, it possesses a surface for indirect prompt injection. The orchestrator mitigates this by delegating research to specialized agents, employing a separate fact-checker agent for provenance verification, and performing judgment-led synthesis to detect and reconcile conflicting information.

interactive-research