Research Loop
Research Loop is a zero-dependency Claude Code plugin that gives your coding agent a complete scientific research workflow. 21 composable skills, 4 specialist subagents, zero dependencies. Clone and go.
How it works
It starts from the moment you open your coding agent and mention anything research-related. As soon as it sees that you're exploring a topic, it doesn't just start searching and dumping information. Instead, it steps back and asks you what you're really trying to figure out.
Once it's teased the research framing out of the conversation, it explores in parallel — papers, repos, debates, open problems — and shows you the landscape in a synthesis short enough to actually read and act on.
After you've picked a direction, your agent finds the gaps, runs them through a Carlini gate (one question at a time, waiting for your answers), and surfaces the ideas worth pursuing. Then it spins up parallel hypothesis lanes, applies gates between them, and kills the weak ones early.
Next up, once you say "go", it launches a subagent-driven experiment loop — proposing code mutations, running benchmarks, annotating results causally, and building a living knowledge graph that remembers what was tried and why it failed.
When you want to understand something deeply, just say "explain X." The learn skill walks you through how experts actually think about a topic — not a summary, but the underlying reasoning structures: core mental models one at a time, the places where the field genuinely disagrees, questions that expose whether you understand or just recognize, and finally a reverse test where you explain it back. Every gap gets logged to the lab notebook.
Need a paper? Say "write a paper on X" and the paper-pipeline skill drives a 14-phase pipeline from topic framing all the way to conference-formatted export.
Need to optimize something? Say "optimize this metric" and the autonomous-iteration skill runs a Karpathy-style Modify → Verify → Keep/Discard loop with automatic rollback.
There's a bunch more to it, but that's the core of the system. And because the skills trigger automatically, you don't need to do anything special. Your agent just has Research Loop.
Sponsorship
If Research Loop has helped you do work that matters and you are so inclined, consider sponsoring the project.
Thanks!
— Alexander
Installation
Research Loop is a zero-dependency Claude Code plugin — it works by dropping skill files into your project. No binaries, no builds, no dependencies.
Plugin install (recommended — no build required)
git clone https://github.com/moralespanitz/research-loop
cd research-loop
That's it. Open a new Claude Code session in this directory and the skills auto-activate.
Plugin install into an existing project
git clone https://github.com/moralespanitz/research-loop /tmp/research-loop
cp -r /tmp/research-loop/.claude /tmp/research-loop/CLAUDE.md /your/project/
Install from the Claude Code marketplace
/plugin install research-loop
Go binary (optional — for CLI users)
A Go binary is available for users who want CLI tools (research-loop init, research-loop start <arxiv-url>, etc.):
go install github.com/moralespanitz/research-loop/cmd/research-loop@latest
research-loop init
Verify installation
Open a new Claude Code session in the workspace and say anything research-related — "I want to explore transformer memory systems" or "explain policy compression." The bootstrap skill loads automatically, then routes to the right sub-skill. You should see the agent announce which skill it's using.
Skills
Research Loop ships 21 skills organized by workflow layer. Skills auto-trigger — you just talk naturally and the right skill activates.
Activation Layer
- bootstrap — Loads on every conversation start. Establishes the skill-loading protocol, Red Flags table to prevent rationalization, and priority rules. The foundation that makes the entire system work.
Core Research Workflow
