skills.rest
epologee
epologeeCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

research

Orchestrate parallel research agents to synthesize findings into atomic improvement plans.

3|1|Updated Apr 4, 2026
Search Tags:#research#agents#synthesis#recursion#knowledge-base#plan-generation#friction-analysis
One-click install
npx skills add https://github.com/epologee/leclause-skills --skill research-epologee
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/epologee/leclause-skills/tree/main/packages/recursion/skills/research
Command: npx skills add https://github.com/epologee/leclause-skills --skill research-epologee

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill automates deep, structured research to surface recurring friction and transform findings into atomic, actionable improvement plans stored locally for review and execution by humans.

Core Features & Use Cases

  • Parallel agent orchestration: Spawns Opus agents for friction analysis and external exploration across multiple rounds to gather diverse evidence.
  • Synthesis and plan generation: Synthesizes results from three research rounds (explore, deepen, contrarian) and writes immutable plan files with a status of proposed.
  • Stateful coordination and safety: Reads and updates state in ~/.claude/recursion, enforces privacy rules, avoids executing plans, and records sources crawled and knowledge base updates.
  • Use Case: Nightly research runs that analyze recent sessions and external sources to propose focused, verifiable improvements to the skill collection.

Quick Start

Invoke the research skill to run the deep research workflow and write proposed atomic plans to ~/.claude/recursion/plans/.

Frequently Asked Questions about research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate friction analysis to generate actionable improvement plans?▼

To automate friction analysis and generate actionable improvement plans, this skill orchestrates parallel agents across three research rounds to synthesize evidence and write atomic, verifiable plans to a local directory.

What is the process for synthesizing research findings into atomic plans?▼

Synthesizing research findings into atomic plans involves running explore, deepen, and contrarian research rounds in parallel, then writing the combined results as immutable proposed files to a local plans directory.

Does this deep research workflow require local read/write access to run?▼

Yes, this deep research workflow requires local read/write access to the ~/.claude/recursion state directory, alongside WebSearch, WebFetch, and agent tooling to orchestrate parallel research rounds.

Can I execute the proposed improvement plans directly within the research skill?▼

No, you cannot execute proposed improvement plans directly within the research skill because it enforces strict privacy and safety rules that prevent executing plans or uploading sensitive data during the orchestration process.

What limitations exist when running nightly research with parallel agents?▼

A key limitation when running nightly research with parallel agents is that the skill only proposes improvements to a local knowledge base; it will not execute the generated plans or upload sensitive session data externally.

Related Skills

Explore More Education & Research Skills→
tkstangtkstang

skeptic

Verify or refute agent claims with evidence and a confidence score.

Community
Advanced
MatthieuszMatthieusz

improve

Analyzes codebases and generates prioritized improvement plans.

Community
Advanced
aivos-xieaivos-xie
1

llm-wiki

Parse Markdown documents into a structured knowledge base for cross-referencing and queries.

Community
Advanced
huyhung9630huyhung9630
1

omc-sciomc

Orchestrate scientific research with parallel analysis, evidence capture, and report synthesis.

Community
Advanced