# strongdm/attractor

nlspec of StrongDM's Attractor, a non-interactive Coding Agent sufficient for use in a Software Factory

Repository: https://github.com/strongdm/attractor
Canonical: https://ross.abutalabs.com/products/attractor
Homepage: https://factory.strongdm.ai/products/attractor
License: Apache-2.0
License Family: permissive
Last push: 2026-03-17T21:33:21+00:00

## Health v2 (maintenance only)
Score: 47/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 72, release rhythm 35, longevity 14
- inputs: {"age_days": 209, "days_push": 169, "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 1277, forks 195 (observed 2026-08-28T04:04:13.306993+00:00)

## What it is
Attractor is a set of Natural Language Specifications (NLSpecs) from StrongDM describing how to build a non-interactive coding agent that composes models, prompts, and tools into a graph-structured pipeline for end-to-end software development. The repository contains human-readable specs (agent loop, unified LLM client, and overall Attractor spec) intended to be directly consumed by coding agents to implement and validate behavior.

## Use cases
- build a non-interactive coding agent from a natural language specification
- automate the software development lifecycle with an LLM-driven graph pipeline
- implement a unified LLM client for multiple model providers
- create a software factory that turns fully specified work into completed code
- orchestrate coding agent phases like implement, optimize, and validate as graph nodes
- evaluate community implementations of an open coding agent spec

## When to choose
- you want a vendor-neutral, open specification for building your own autonomous coding agent
- you need a deterministic, resumable, graph-structured agent pipeline for fully specified tasks
- you want to feed a spec directly to a coding agent like Claude Code or Codex to generate an implementation

## When to avoid
- you need a ready-to-run coding agent binary rather than specifications to implement
- your tasks are not fully specified and require interactive human-in-the-loop coding
- you want a framework with shipped code and APIs instead of natural language documents

## Facets
- artifact type: learning-resource
- maturity: active
- function: agent-framework, llm-inference, prompt-engineering, developer-tools, workflow-automation
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: cross-platform, cli
- tags: nl-spec, coding-agent, software-factory, specification, agentic-loop, non-interactive-agent, graph-pipeline, sdlc-automation, ai-agents, automation

## Member repositories
- strongdm/attractor (main) score 47

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:13.306993+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 2026-08-30T05:02:52.037213+00:00, confidence not recorded.
  - readme: https://github.com/strongdm/attractor (fetched 2026-08-28T04:04:13.306993+00:00, sha 0b5f8d1b8ed4)
  - homepage: https://factory.strongdm.ai/products/attractor (fetched 2026-08-29T12:14:10.349582+00:00, sha 612fbf89ffbc)
- Data as of 2026-08-30T08:39:29.467469+00:00.