Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
A research agent that conducts structured investigations on tech and internet topics. It runs a three-phase workflow: broad search, focused iterative deep dives, and final report synthesis with inline citations. Treats user-provided terms as authoritative, never substituting or correcting them against its internal knowledge.
Built-in stalemate handling stops wasted effort after two failed attempts per sub-goal, and a hard word-count cap keeps reports focused.
When to use it
- Researching a newly released AI model before it appears in common knowledge bases
- Comparing competing cloud platforms across multiple authoritative sources
- Tracing version history and key changes of an open-source framework
- Producing a cited technical report on an emerging internet protocol
- Investigating a specific product or company across news, papers, and community sources
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Research the latest release of Claude 4 Sonnet
- 1Formulates 3-5 search angles around the exact term 'Claude 4 Sonnet' without substitution or correction
- 2Creates a research log wiki document and records all search result titles, abstracts, and URLs
- 3Reads up to 4 authoritative URLs in parallel and appends key facts and figures to the log
- 4Identifies the most important unresolved question and runs 1-2 focused follow-up searches
- 5Synthesizes all log data into a final report with inline citations and a numbered reference list
A structured research report covering background, core findings, and conclusion, with every key claim linked to a source URL