andrehuang

andrehuang / research-companion

Public

Strategic research thinking agents for Claude Code — idea evaluation, project triage, and structured brainstorming. Helps you decide which papers to write, not just how to write them.

530
46
100% credibility
Found Apr 08, 2026 at 534 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

This repository provides AI agents and structured sessions for Claude Code and Codex to help researchers critically evaluate ideas, strategize projects, and brainstorm effectively before committing time.

How It Works

1
🔍 Discover Research Companion

You hear about this smart helper that acts like a trusted research advisor, guiding you on which ideas are worth pursuing.

2
âž• Add to Your AI Chat

You simply connect it to your AI conversation tool, like Claude or Codex, with a quick addition.

3
đź’­ Share Your Research Idea

You describe your idea or project challenge, and the advisor jumps in to think deeply with you.

4
⚖️ Get Honest Critique

It evaluates your idea across key angles like novelty, impact, and risks, giving a clear pursue, refine, or drop verdict.

5
đź’ˇ Brainstorm Fresh Angles

Together, you explore cross-field connections and new framings to spark truly original directions.

6
đź“‹ Run Full Session

Launch a guided journey from seeding ideas to framing conclusions and making the final call.

âś… Decide Confidently

You walk away with saved evaluations, a watch list, and focus on high-impact research that matters.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 534 to 530 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is research-companion?

Research-companion delivers AI agents as plugins for Claude Code and OpenAI Codex, acting as your strategic research companion—evaluating ideas, triaging projects, and running structured brainstorming sessions. It tackles the gap before writing: killing bad ideas early via 7 evaluation dimensions like novelty and competitive landscape, or guiding 6-phase ideation from seed to decision. Like a research companion book crossed with strategic research insights, it uses YAML configs and marketplace installs for commands like `/research-companion` to persist evaluations across sessions.

Why is it gaining traction?

It stands out by focusing on "which research to pursue" over drafting, with opinionated agents that deliver blunt verdicts—Pursue/Refine/Kill—saving months on doomed projects. Features like cross-field brainstorming, conclusion-first tests, and watch lists for scooping risks hook users tired of generic AI writing tools. With 530 stars, developers grab it for its Claude/Codex compatibility and persistent markdown outputs that build a research journal over time.

Who should use this?

PhD students and ML researchers triaging project backlogs, academics stress-testing grant ideas, or indie researchers spotting cross-domain nuggets in privacy or synthesis. It's for anyone in strategic research associates roles needing a devil's advocate before committing cycles, not post-hoc polishers.

Verdict

Grab it if you're in research—solid docs and user-friendly installs make the 530 stars earned, despite the 1.0% credibility score signaling early maturity. Test on your next idea; it'll sharpen your strategic research fund decisions fast.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.