Agent skill · research science · code-yeongyu
ulw-research
Runs maximum-saturation research with a cooperating team, claim-graph gating, and a cited, QA'd deliverable. Use when the user explicitly asks for research or a deep investigation, including any 'ulw' research wording.
Why this skill is useful
Adds a comprehensive research orchestration framework with live journaling, empirical verification, and a synthesis process that significantly enhances the AI's research capabilities.
What it needs
About 12k tokens when loaded. Last updated 2026-08-07. 67,387 stars on the source repository.
What this skill does
Codex Harness Tool Compatibility This skill may include examples copied from the OpenCode harness. In Codex, do not call OpenCode-only tools such as callomoagent(...), task(...), backgroundoutput(...), or team(...) literally. Translate those examples to Codex native tools: OpenCode example Codex tool to use --- --- callomoagent(subagenttype="explore", ...) multiagentv1.spawnagent({"message":"TASK: act as an explorer. ...","agenttype":"explorer","forkcontext":false}) callomoagent(subagenttype="librarian", ...) multiagentv1.spawnagent({"message":"TASK: act as a librarian. ...","agenttype":"librarian","forkcontext":false}) task(subagenttype="plan", ...) multiagentv1.spawnagent({"message":"TASK: act as a planning agent. ...","agenttype":"plan","forkcontext":false}) task(subagenttype="oracle", ...) for final verification multiagentv1.spawnagent({"message":"TASK: act as a rigorous reviewer. ...","agenttype":"lazycodex-gate-reviewer","forkcontext":false}) task(category="...", ...) for implementation or QA multiagentv1.spawnagent({"message":"TASK: act as an implementation or QA worker. ...","forkcontext":false}) backgroundoutput(taskid="...") multiagentv1.waitagent(...) for mailbox signals team(...) Use Codex native subagents via multiagentv1.spawnagent and multiagentv1.waitagent; use multiagentv1.sendinput and multiagentv1.closeagent only when exposed in the active tools list Role-specific behavior must be described in a self-contained message. Use forkcontext: false to start the child with only the initial prompt (no parent history); use forkcontext: true only when full parent history is truly required. Include any required conversation context, files, diffs, constraints, and requested skill names directly in the spawned agent's message. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills code-yeongyu/ulw-research--a4d7cf