# evalstate/fast-agent

Code, Build and Evaluate agents - excellent Model and Skills/MCP/ACP/A2A Support

Repository: https://github.com/evalstate/fast-agent
Canonical: https://ross.abutalabs.com/products/fast-agent
Homepage: https://fast-agent.ai
Language: Python
License: Apache-2.0
License Family: permissive
Topics: agent, agent-framework, agent-skills, cli, mcp, mcp-client, mcp-server, python, skills, tui, acp, a2a
Last push: 2026-08-26T22:12:14+00:00

## Health v2 (maintenance only)
Score: 83/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 42
- inputs: {"age_days": 592, "days_push": 7, "days_rel": 10, "gap_med": 3.5, "n_releases_24m": 105}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3902, forks 437 (observed 2026-08-28T04:08:28.595109+00:00)

## What it is
fast-agent is a Python CLI-first framework for building, running, and evaluating LLM agents with first-class MCP, ACP, and A2A protocol support. It provides an interactive terminal UI with shell integration, skills/plugins, workflow orchestration, and broad provider support including local llama.cpp models.

## Use cases
- build and evaluate llm agents from the terminal
- coding agent with shell and subagent support
- connect and test mcp servers with sampling and elicitation
- expose an agent as an a2a or acp endpoint
- run local llama.cpp models with an agent harness
- evaluate model accuracy and cost across providers
- manage agent skills and plugins

## When to choose
- you want a terminal-native, scriptable agent harness with strong MCP support
- you need ACP/A2A client and server interoperability
- you want broad provider coverage including local models and Codex subscriptions
- you need agent evaluation and workflow tooling in Python

## When to avoid
- you need a hosted GUI or web-based agent builder
- you only want a simple chatbot SDK without protocol integrations
- you require a non-Python runtime

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, llm-inference, cli, mcp, chatbot, developer-tools, terminal-ui, plugin-system
- domain: large-language-models, developer-tools
- platform: python, cli, cross-platform, windows
- tags: mcp-client, mcp-server, acp, a2a, agent-skills, coding-agent, tui, llm-evaluation, card-packs, ai-agents, command-line, automation, macos, linux

## Member repositories
- evalstate/fast-agent (main) score 83

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:28.595109+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 2026-08-29T18:25:10.121223+00:00, confidence not recorded.
  - readme: https://github.com/evalstate/fast-agent (fetched 2026-08-28T04:08:28.595109+00:00, sha 9803cc22977e)
  - homepage: https://fast-agent.ai (fetched 2026-08-29T09:19:21.919679+00:00, sha 911a5312b404)
  - site_page: https://fast-agent.ai/a2a/getting-started (fetched 2026-08-29T09:19:21.928745+00:00, sha c5620a961924)
  - site_page: https://fast-agent.ai/ref/generated_docs (fetched 2026-08-29T09:19:21.930811+00:00, sha d4a0fc25e8e7)
  - site_page: https://fast-agent.ai/ref/docs_automation (fetched 2026-08-29T09:19:21.932475+00:00, sha 56b5b63730ed)
  - site_page: https://fast-agent.ai/ref/about (fetched 2026-08-29T09:19:21.934417+00:00, sha 5b81ce0c5763)
  - site_page: https://fast-agent.ai/models (fetched 2026-08-29T09:19:21.936315+00:00, sha 1dfc7b563c36)
  - site_page: https://fast-agent.ai/agents/plugins (fetched 2026-08-29T09:19:21.939199+00:00, sha a9065d7654fd)
- Data as of 2026-08-30T08:39:29.467469+00:00.