# apache/burr

Build applications that make decisions (chatbots, agents, simulations, etc...). Monitor, trace, persist, and execute on your own infrastructure.

Repository: https://github.com/apache/burr
Canonical: https://ross.abutalabs.com/products/burr
Homepage: https://burr.apache.org/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: burr, dags, graphs, llmops, llms, mlops, persistent-data-structure, state-machine, state-management, visibility, ai, chatbot-framework, generative-ai, hacktoberfest
Last push: 2026-08-25T21:51:19+00:00

## Health v2 (maintenance only)
Score: 87/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 83, longevity 67
- inputs: {"age_days": 947, "days_push": 8, "days_rel": 115, "gap_med": 7, "n_releases_24m": 28}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2528, forks 188 (observed 2026-08-28T04:06:58.706770+00:00)

## What it is
Apache Burr is a Python framework for building stateful applications that make decisions, such as chatbots, agents, and simulations, using actions and transitions as simple Python building blocks. It includes a self-hosted UI for real-time monitoring, tracing, and debugging, plus pluggable state persistence and human-in-the-loop support.

## Use cases
- build a chatbot with persistent conversation state
- create multi-agent LLM workflows with branching and parallelism
- trace and debug LLM application steps in real time
- add human-in-the-loop approval steps to an agent
- persist and resume long-running AI application state
- build decision-making simulations in Python
- monitor production LLM apps on my own infrastructure

## When to choose
- you want a pure-Python, no-DSL way to build stateful LLM agents or chatbots
- you need built-in observability, tracing, and state persistence you can self-host
- you want human-in-the-loop pauses and replay/testing of AI workflows

## When to avoid
- you need a fully managed cloud agent platform rather than self-hosted infrastructure
- you want a high-level opinionated agent framework with prebuilt tools rather than composable primitives
- you are not building stateful or decision-driven applications

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, state-management, workflow-automation, chatbot, monitoring, tracing, rag, llm-inference
- domain: large-language-models, chatbots, developer-tools, artificial-intelligence
- platform: python, cross-platform, self-hosted
- tags: state-machine, llmops, observability, human-in-the-loop, dags, persistence, apache-incubating, ai-agents, retrieval-augmented-generation

## Member repositories
- apache/burr (main) score 87

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:58.706770+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 2026-08-30T02:25:23.031947+00:00, confidence not recorded.
  - readme: https://github.com/apache/burr (fetched 2026-08-28T04:06:58.706770+00:00, sha 94f5ef9088cf)
  - homepage: https://burr.apache.org/ (fetched 2026-08-29T10:07:47.669754+00:00, sha cc83dadb1e92)
  - site_page: https://burr.apache.org/docs (fetched 2026-08-29T10:07:47.679957+00:00, sha 38fa20fc13f8)
  - registry_pypi: https://pypi.org/pypi/burr/json (fetched 2026-08-29T10:07:47.682772+00:00, sha 92b70f9855ff)
- Data as of 2026-08-30T08:39:29.467469+00:00.