# The-Pocket/PocketFlow

Pocket Flow: 100-line LLM framework. Let Agents build Agents!

Repository: https://github.com/The-Pocket/PocketFlow
Canonical: https://ross.abutalabs.com/products/pocketflow
Homepage: https://the-pocket.github.io/PocketFlow/
Language: Python
License: MIT
License Family: permissive
Topics: agents, flow-based-programming, flow-engineering, llm-agent, llm-framework, agentic-workflow, agentic-ai, agentic-framework, aiagent, aiagents, pocket-flow, workflow, workflow-orchestration, ai-framework, ai-frameworks, artificial-intelligence, retrieval-augmented-generation, large-language-model, large-language-models, pocketflow
Last push: 2026-07-26T16:13:20+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 94, release rhythm 8, longevity 44
- inputs: {"age_days": 617, "days_push": 38, "days_rel": 526, "gap_med": null, "n_releases_24m": 1}
- flags: prerelease_only
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 11135, forks 1215 (observed 2026-08-28T04:10:46.187855+00:00)

## What it is
Pocket Flow is a minimalist 100-line LLM framework in Python with zero dependencies, modeling LLM applications as a graph of nodes connected by actions with a shared store. It supports agents, workflows, RAG, multi-agent coordination, and batch/async patterns while leaving vendor-specific utilities to the user.

## Use cases
- build llm agents in python without heavy frameworks
- orchestrate multi-step llm workflows as a graph
- implement rag pipelines with a minimal framework
- coordinate multiple ai agents
- let ai coding assistants generate agent code
- avoid vendor lock-in when calling llm apis

## When to choose
- you want a tiny, zero-dependency LLM framework you can read and copy entirely
- you need agents, workflows, or RAG without vendor lock-in
- you plan to use AI coding agents to build LLM applications
- you want to implement your own LLM wrappers and utilities

## When to avoid
- you need built-in integrations for LLM providers, vector databases, or web search
- you want production-grade features like observability, retries, and tooling out of the box
- you prefer batteries-included frameworks like LangChain

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, rag, workflow-automation, llm-inference, prompt-engineering
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, cross-platform
- tags: minimalist, zero-dependency, graph-based, flow-based-programming, agentic-coding, multi-agent, llm-orchestration, mit-license, ai-agents, retrieval-augmented-generation

## Member repositories
- The-Pocket/PocketFlow (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:46.187855+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 2026-08-29T17:16:46.033619+00:00, confidence not recorded.
  - readme: https://github.com/The-Pocket/PocketFlow (fetched 2026-08-28T04:10:46.187855+00:00, sha 8ae40f8daa9e)
  - homepage: https://the-pocket.github.io/PocketFlow/ (fetched 2026-08-29T08:15:17.972489+00:00, sha 23e81ac9613f)
  - registry_pypi: https://pypi.org/pypi/pocketflow/json (fetched 2026-08-29T08:15:17.981542+00:00, sha 201df6de95c5)
- Data as of 2026-08-30T08:39:29.467469+00:00.