# humanlayer/12-factor-agents

What are the principles we can use to build LLM-powered software that is actually good enough to put in the hands of production customers?

Repository: https://github.com/humanlayer/12-factor-agents
Canonical: https://ross.abutalabs.com/products/12-factor-agents
Language: TypeScript
License: NOASSERTION
License Family: other
Topics: agents, ai, context-window, framework, llms, memory, orchestration, prompt-engineering, rag, 12-factor, 12-factor-agents
Last push: 2025-09-21T14:37:40+00:00

## Health v2 (maintenance only)
Score: 39/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 43, release rhythm 35, longevity 37
- inputs: {"age_days": 521, "days_push": 346, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 25510, forks 1935 (observed 2026-08-28T04:11:38.200116+00:00)

## What it is
A guide of twelve principles for building reliable, production-grade LLM-powered agent applications, inspired by 12 Factor Apps. It includes explanatory content, examples, and a scaffolding tool (create-12-factor-agent) written in TypeScript.

## Use cases
- learn how to build production-ready LLM agents
- principles for designing reliable AI agent software
- understand context window management for LLM apps
- decide whether to use an agent framework or roll my own stack
- best practices for putting LLM features in front of customers
- context engineering techniques for agents

## When to choose
- you are designing or reviewing an LLM-powered product architecture
- you want framework-agnostic principles rather than a specific library
- you are evaluating agent frameworks and want to know why teams roll their own

## When to avoid
- you need a ready-made runtime framework or SDK rather than guidance
- you want a turnkey agent product with no implementation work
- you need guaranteed long-term maintenance, as it is primarily an evolving methodology document

## Facets
- artifact type: learning-resource
- maturity: active
- function: agent-framework, prompt-engineering, rag, llm-inference
- domain: large-language-models, developer-tools, tutorials
- platform: cross-platform
- tags: 12-factor, context-engineering, llm-applications, production-ai, agent-design-principles, methodology, ai-agents, nodejs, typescript

## Member repositories
- humanlayer/12-factor-agents (main) score 39

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:38.200116+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 2026-08-29T16:56:02.745688+00:00, confidence not recorded.
  - readme: https://github.com/humanlayer/12-factor-agents (fetched 2026-08-28T04:11:38.200116+00:00, sha fb598d6dcf2a)
- Data as of 2026-08-30T08:39:29.467469+00:00.