# jennyzzt/dgm

Darwin Gödel Machine: Open-Ended Evolution of Self-Improving Agents

Repository: https://github.com/jennyzzt/dgm
Canonical: https://ross.abutalabs.com/products/dgm
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2025-08-13T10:40:14+00:00

## Health v2 (maintenance only)
Score: 35/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 36, release rhythm 35, longevity 33
- inputs: {"age_days": 467, "days_push": 385, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2254, forks 441 (observed 2026-08-28T04:06:31.150065+00:00)

## What it is
Darwin Gödel Machine (DGM) is a research implementation of a self-improving AI agent that iteratively rewrites its own codebase and validates each modification against coding benchmarks like SWE-bench and Polyglot. It uses foundation models (OpenAI/Anthropic APIs) within a Docker-based evaluation loop to evolve agents via open-ended evolution.

## Use cases
- run self-improving coding agent experiments
- evolve an agent that modifies its own code
- benchmark coding agents on SWE-bench
- research open-ended evolution for AI agents
- reproduce the Darwin Gödel Machine paper
- evaluate LLM agents on Polyglot benchmark

## When to choose
- you want to reproduce or extend the DGM self-improving agent research
- you need an experimental framework for evolving coding agents against benchmarks
- you're studying open-ended evolution and self-modification in LLM agents

## When to avoid
- you need a production-ready coding assistant
- you can't provide GPU/API access or Docker infrastructure
- you want a stable, supported library rather than research code

## Facets
- artifact type: library
- maturity: experimental
- function: agent-framework, llm-inference, benchmarking, machine-learning
- domain: artificial-intelligence, large-language-models
- platform: python
- tags: self-improving-agents, open-ended-evolution, swe-bench, darwin-godel-machine, research-code, autonomous-coding-agents, ai-agents, research, docker, linux, macos

## Member repositories
- jennyzzt/dgm (main) score 35

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:31.150065+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 2026-08-30T02:43:46.496875+00:00, confidence not recorded.
  - readme: https://github.com/jennyzzt/dgm (fetched 2026-08-28T04:06:31.150065+00:00, sha de87cfbf4f32)
- Data as of 2026-08-30T08:39:29.467469+00:00.