Produce verified structured research outputs across multi-agent investigation workflows.
npx skills add https://github.com/davdunc/pai-framework --skill research-davdunc
Please help me install this Agent Skill. Skill: Research Source: https://github.com/davdunc/pai-framework/tree/main/skills/Research Command: npx skills add https://github.com/davdunc/pai-framework --skill research-davdunc
Research turns ambiguous questions into structured findings—covering quick lookups, standard multi-agent research, extensive parallel exploration, deep iterative investigations, and high-signal content extraction—while enforcing URL verification to prevent hallucinated links.
Use the Research skill when you need verified information by saying: “Do research on [your topic]” (or “quick research on [your topic]”).
High-intent search queries and answers about installing and using this skill.
Multi-agent web research prevents hallucinated links by enforcing mandatory URL verification during information synthesis, ensuring all extracted content and references point to actual sources. This verification step is built into every research workflow.
To run a deep investigation on a broad topic landscape, trigger the workflow by requesting research on your topic. The system iteratively executes landscape, entities, and deep-dive analysis phases, building a structured vault of verified findings.
Yes, you can extract and synthesize information from a YouTube URL. The research workflow includes content retrieval and scraping layers specifically designed for YouTube content extraction, yielding high-signal insights from video sources.
For a quick lookup, use the fast single-agent mode by specifying quick research. For extensive parallel exploration, trigger the extensive workflow to deploy a 12-agent configuration for comprehensive information gathering and synthesis.
Yes, the information synthesis workflow supports custom configurations. It optionally loads customizations from SKILLCUSTOMIZATIONS before executing tasks, allowing you to tailor the research and extraction parameters to your specific needs.
ValyntPerforms end-to-end load testing and tunes HPA autoscaling for ValueOS agent workloads using K6 and Locust.
otto-poblyshEnforce root-cause investigation before fixing bugs and test failures.
handsomelong922Log learnings and corrections to .learnings markdown files for future sessions.
landonking-gifSearch the web, extract data points, and prepare structured report content.
