# wanxingai/LightAgent

LightAgent: Lightweight Python framework for OpenAI-compatible agents with tools, memory, guardrails, tracing, lifecycle hooks, multi-agent collaboration, and workflows.

Repository: https://github.com/wanxingai/LightAgent
Canonical: https://ross.abutalabs.com/products/lightagent
Language: Python
License: Apache-2.0
License Family: permissive
Topics: agent-framework, agents, ai-agent, llm, mcp, multi-agent, openai, openai-api, python, tool-calling, agent-hooks, lifecycle-hooks
Last push: 2026-08-21T14:17:39+00:00

## Health v2 (maintenance only)
Score: 87/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 98, longevity 42
- inputs: {"age_days": 590, "days_push": 12, "days_rel": 18, "gap_med": 14, "n_releases_24m": 30}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1211, forks 171 (observed 2026-08-28T04:04:00.147760+00:00)

## What it is
LightAgent is an ultra-lightweight open-source Python framework for building LLM-powered agents with tools, persistent memory, guardrails, tracing, and tree-of-thought reasoning. It supports multi-agent collaboration, MCP integration over stdio/SSE, and exposes OpenAI-compatible streaming APIs that work with any modern LLM such as OpenAI, DeepSeek, and Qwen.

## Use cases
- build a lightweight ai agent with tool calling in python
- create multi-agent collaboration workflows with llms
- add memory and guardrails to an openai-compatible chatbot
- connect agents to mcp servers over stdio or sse
- serve an openai-compatible streaming agent api for a chat interface
- build self-learning agents with lifecycle hooks and tracing

## When to choose
- you want a minimal, fast-to-set-up agent framework in pure Python
- you need OpenAI-compatible streaming output to drop into existing chat UIs
- you want built-in memory, guardrails, tracing, and MCP support without heavy dependencies
- you work with multiple LLM providers like OpenAI, DeepSeek, or Qwen

## When to avoid
- you need a mature, battle-tested enterprise agent platform with large ecosystem support
- you require deep integrations with a specific vendor framework like LangChain or AutoGen
- you need non-Python languages or heavy distributed multi-agent orchestration at scale

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, llm-inference, mcp, chatbot, workflow-automation
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, cross-platform
- tags: tool-calling, multi-agent, openai-compatible, tree-of-thought, memory, guardrails, tracing, lifecycle-hooks, self-hosted, ai-agents

## Member repositories
- wanxingai/LightAgent (main) score 87

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:00.147760+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 2026-08-30T06:18:05.821991+00:00, confidence not recorded.
  - readme: https://github.com/wanxingai/LightAgent (fetched 2026-08-28T04:04:00.147760+00:00, sha 9b331d281ac7)
  - registry_pypi: https://pypi.org/pypi/lightagent/json (fetched 2026-08-29T12:26:11.132588+00:00, sha c3c8b3b57d75)
- Data as of 2026-08-30T08:39:29.467469+00:00.