# skyllwt/AutoSci

Karpathy's LLM-Wiki vision, fully realized — wiki-centric full-lifecycle AI research platform powered by Claude Code

Repository: https://github.com/skyllwt/AutoSci
Canonical: https://ross.abutalabs.com/products/autosci
Language: Python
License: MIT
License Family: permissive
Last push: 2026-08-24T08:59:54+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 10
- inputs: {"age_days": 146, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1649, forks 210 (observed 2026-08-28T04:05:16.752413+00:00)

## What it is
AutoSci is a wiki-centric AI research platform that automates the full scientific research lifecycle — reading literature, thinking, running experiments, writing, and evolving — powered by Claude Code (with Codex and OpenCode previews). It implements Karpathy's LLM-Wiki vision with memory that compounds across projects via components like SciMem, SciFlow, SciDAG, and SciEvolve.

## Use cases
- automate literature reviews on a research topic
- generate a wiki of knowledge from arxiv papers
- run an AI agent that reads papers and writes research notes
- manage a full AI-assisted scientific research lifecycle
- build compounding research memory across projects
- get daily arxiv paper recommendations by email
- use claude code as a research assistant

## When to choose
- you want an AI agent to automate reading, synthesizing, and writing research content
- you already use Claude Code and want a research-oriented workflow on top of it
- you need a wiki-style knowledge base that grows with each research project

## When to avoid
- you need a fully stable, production-hardened tool — it is in internal beta
- you don't use Claude Code, Codex, or OpenCode as your coding agent runtime
- you need offline research without LLM API access

## Facets
- artifact type: application
- maturity: experimental
- function: agent-framework, llm-inference, rag, workflow-automation, documentation, search-engine
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, cli, cross-platform
- tags: ai-research-agent, claude-code, wiki, scientific-research, llm-agent, knowledge-management, arxiv, codex, opencode, ai-agents, research, natural-language-processing

## Member repositories
- skyllwt/AutoSci (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:16.752413+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:45:10.565618+00:00, confidence not recorded.
  - readme: https://github.com/skyllwt/AutoSci (fetched 2026-08-28T04:05:16.752413+00:00, sha f0ae2857362b)
- Data as of 2026-08-30T08:39:29.467469+00:00.