AutoResearchClaw

(β˜… 14,055)

Fully autonomous & self-evolving research from idea to paper. Chat an Idea. Get a Paper. 🦞

Reference Links

πŸ€— Hugging Face MiniMax platform MiniMax China platform FARS

Owner Repos

SimpleMem β˜… 3,708

SimpleMem: Efficient Lifelong Memory for LLM Agents β€” Text & Multimodal

Python
MetaClaw β˜… 3,492

🦞 Just talk to your agent β€” it learns and EVOLVES 🧬.

Python
Agent0 β˜… 1,252

[ICML'26 & COLM'26] Agent0 Series: Self-Evolving Agents from Zero Data

Python
SkillRL β˜… 944

SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning

Python
AutoHarness β˜… 368

AutoHarness: Automated Harness Engineering for AI Agents

Python
MDocAgent β˜… 354

MDocAgent: A Multi-Modal Multi-Agent Framework for Document Understanding

Python

File Explorer

ZIP Download
  • .gitignore
  • config.researchclaw.example.yaml
  • CONTRIBUTING.md
  • LICENSE
  • prompts.default.yaml
  • pyproject.toml
  • README.md
  • RESEARCHCLAW_AGENTS.md
  • RESEARCHCLAW_CLAUDE.md
  • run_hep_pipeline.sh
  • sentinel.sh

# Installation Guide

python
Prerequisites

Setup Steps

pip install -e .

Install the project in editable mode.

cp config.researchclaw.example.yaml config.researchclaw.yaml

Copy the example configuration file to create the local config.

researchclaw setup

Run initial setup for AutoResearchClaw.

researchclaw init

Initialize required workspace and resources.

researchclaw run --topic "Your research idea here" --auto-approve

Run the pipeline fully autonomously based on the research topic.

Key Commands

pip install -e .

Install the package from the current directory in editable mode.

researchclaw setup

Set up the environment for AutoResearchClaw.

researchclaw init

Initialize project settings and local database.

researchclaw run --topic "Quantum Error Correction" --auto-approve

Start the autonomous research pipeline to generate a paper.

You must configure your LLM API keys in config.researchclaw.yaml before running the pipeline.

# Use via CDN

jsDelivr

jsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.

Link
Example

Release Timeline

Actively Maintained

Command Glossary

Commands referenced in this DOCs, explained below.

πŸ”

claude

View Details β–Ό

An agent-based coding tool that understands your code base and helps you code faster through natural language commands.

claude prompt

Execute with prompt:

claude update

Update `claude`:

claude mcp list

Get the list of specified MCP servers:

πŸ”

codex

View Details β–Ό

Natural language code assistant for the terminal, powered by OpenAI.
Reads and edits files in your current directory to fulfill requests.

codex

Start an interactive Codex session in the current directory:

codex "{{prompt}}"

Run a single Codex command using a prompt:

codex --sandbox workspace-write "{{prompt}}"

Run a prompt allowing Codex to edit files in the workspace:

πŸ”

gemini

View Details β–Ό

Launch an interactive prompt with Gemini AI.

gemini

Start a REPL session to chat interactively:

{{echo "Summarize the history of Rome"}} | gemini {{[-p|--prompt]}} -

Send the output of another command to Gemini and exit immediately:

gemini {{[-m|--model]}} {{model_name}}

Use a specific model (default: auto-gemini-3):

πŸ”

git clone

View Details β–Ό

Clone an existing repository.

git clone {{remote_repository_location}} {{path/to/directory}}

Clone an existing repository into a new directory (the default directory is the repository name):

git clone --recursive {{remote_repository_location}}

Clone an existing repository and its submodules:

git clone {{[-n|--no-checkout]}} {{remote_repository_location}}

Clone only the `.git` directory of an existing repository:

πŸ”

opencode

View Details β–Ό

An AI coding agent.
Some subcommands such as `auth`, `models`, `web`, etc. have their own usage documentation.

opencode

Start the interactive TUI:

opencode {{[-c|--continue]}}

Continue the most recent session:

opencode run "{{message}}"

Run opencode in non-interactive mode by passing a prompt directly:

πŸ”

pip install

View Details β–Ό

Install Python packages.

pip install {{package1 package2 ...}}

Install one or more packages:

pip install {{package1 package2 ...}} {{[-U|--upgrade]}}

Upgrade all specified packages to the latest version, installing any that are not already present:

pip install {{package}}=={{version}}

Install a specific version of a package:

πŸ”

python

View Details β–Ό

Python language interpreter.

python

Start a REPL (interactive shell):

python {{path/to/file.py}}

Execute a specific Python file:

python -i {{path/to/file.py}}

Execute a specific Python file and start a REPL:

πŸ”

python3

View Details β–Ό

This command is an alias of `python`.

tldr python

View documentation for the original command:

// repository documentation