Skip to content
Get packs

Research Refine Pipeline

Chains method refinement and experiment planning into one research proposal and roadmap

Source Download skill
Repository
wanshuiyin/Auto-claude-code-research-in-sleep
Stars
18,482
Licence
MIT
Files
1 file

What it does

This skill makes an AI agent run a full research-planning workflow in one go. It starts from a vague research idea, refines it into a focused method proposal, reviews it, then writes a detailed experiment plan, tracker and short summary of what to run next. It delegates to two sibling skills (research-refine and experiment-plan) and can call reviewer models through a configurable reviewer route, which by default is Codex via MCP.

Great for

  • Researchers who want a complete method plus experiment roadmap from a rough idea
  • Students drafting a paper plan who want structured review history and claims
  • Teams already using the sibling research-refine and experiment-plan skills who want them chained

Skip it if

  • Simple one-off questions: the workflow writes several files and runs multi-stage review, which is heavy for small tasks
  • Anyone expecting finished experiments: it only plans, and the summary points to a separate /run-experiment step
  • Projects without the sibling skills and reviewer setup, since the instructions depend on them and on shared protocol files not included here

Try asking

  • “Turn my rough idea on sparse attention for long videos into a final proposal and experiment roadmap”
  • “Run the research refine pipeline end-to-end on my draft method and tell me what to run first”
  • “Generate both the method and experiment plan together for my workshop submission”
Step-by-step guide: how to use Research Refine PipelineInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    research-refine-pipeline.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip research-refine-pipeline.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip research-refine-pipeline.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Start with a clear problem statement and target venue so the proposal stays anchored and the plan is not generic
  • Check refine-logs/ outputs and MANIFEST.md after the run, since the agent writes many files into your project
  • Review reviewer-route settings before running, because reviewer calls may send your research text to an external model service

What's inside

  • SKILL.md

Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.