AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Dev Tooling/
fast-agent
fast-agent logo

fast-agent

Active·★ 4.0k·Apache-2.0·Updated 2026-11-18
★ Trending

fast-agent is a Python framework for building and chatting with multimodal agents and workflows from the command line. It describes itself as the first framework with complete, end-to-end tested MCP feature support, including sampling and elicitations. Agents are declared with a decorator, for example `@fast.agent(instruction=...)`, then you send messages in code or start an interactive chat; the README also covers combining agents with MCP servers, agents as tools, a workflow called MAKER, MCP OAuth with keyring storage, and an interactive shell.

Install it with `uv pip install fast-agent-mcp`, then `fast-agent go` starts an interactive session, optionally with a remote MCP server (`--url`) or a local Ollama model, and `fast-agent scaffold` generates an example agent and config. `uv run agent.py --transport http --port 8001` exposes an agent as an MCP server. Native providers are Anthropic, OpenAI and Google, plus Azure, Ollama, Deepseek and many more through TensorZero, with structured outputs, PDF and vision support. Recent features listed include Agent Skills (SKILL.md), MCP-UI and shell mode. The project builds on `mcp-agent` and is Apache-2.0 licensed.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-11-15.

License
Apache-2.0
Latest release
v0.10.53 · 2026-11-05
Last commit
2026-11-15
Contributors
89
Forks
469
Open issues and PRs
52
First published
2025-01-18

Source: GitHub, npm and PyPI public APIs.

#acp#agent#agent-framework#agent-skills#cli#mcp#mcp-client#mcp-server
$ Install
$ uv pip install fast-agent-mcp
↗ Visit site★ GitHub
01

Features

01Sophisticated multimodal Agent and Workflow creation.
02Comprehensive LLM provider support (e.g., Anthropic, OpenAI, Google, Ollama).
03Structured outputs, PDF, and Vision support.
04CLI-first with interactive terminal UI (TUI-style) and live streaming responses.
05Advanced agentic workflows like Chain, Parallel, MAKER, and Agents As Tools.
02

Why choose it

+Declarative decorator syntax: a working agent is a few lines of Python
+Broad MCP feature coverage, including sampling, elicitation, OAuth and transport diagnostics
+Native support for Anthropic, OpenAI and Google, plus local models through Ollama
+Can expose an agent as an MCP server with one flag
03

Trade-offs

!The README leans on the project's own claims (for example being the first or only tool with certain features); verify against your needs
!Documentation lives in a submodule and a separate site, so you need `--recurse-submodules` when cloning
!Python-only and CLI-first; there is no full-screen GUI
!Contribution guidelines and roadmap are marked as coming soon
04

Maintenance check

ActiveChecked: 2026-11-06
Last commit
2026-11-06
Commits, last 90 days
100+
Releases, last 90 days
11
Latest release
v0.10.53 (2026-11-05)
Open issues
43
Recent issues with at least one comment
1/5
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 8, 8, 11, 8, 4, 22, 11, 36, 16, 14, 11, 21

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #522 Future ideas: True REPL, First-class multimodal generation, more open-source models, internal MBUS, notebook client, massive parallelism7 comments · 2025-11-27
  • #394 integrate structured output into chains?7 comments · 2025-09-16
  • #244 Add parameters max_context_length_per_message and max_total_context_length3 comments · 2025-06-20
  • #168 A very small suggestion, do we need a fast-agent logo?3 comments · 2025-05-14
  • #598 discussion regarding userfriendly spec violation message2 comments · 2026-01-05

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

Anthropic
Native
Verified via docs
OpenAI
Native
Verified via docs
Google
Native
Verified via docs
Azure
Supported
Verified via docs
Ollama
Supported
Verified via docs
Deepseek
Supported
Verified via docs
06

Quick start

1
$ uv pip install fast-agent-mcp
07

Use cases

↳Rapid development and testing of sophisticated multimodal AI agents and workflows.
↳Diagnosing and tuning agent applications, including those requiring human input.
↳Implementing advanced agentic patterns for error reduction or complex task decomposition.
08

Alternatives

⚡Top VS:
VSgradle-mcp★68→VSeclipse-chatgpt-plugin★181→VSturbomcp★99→VSmcp★78→
gradle-mcp logo
gradle-mcp★ 68
A comprehensive MCP server for Gradle.
VSCompare→
Embody logo
Embody★ 222
Have a conversation with TouchDesigner
eclipse-chatgpt-plugin logo
eclipse-chatgpt-plugin★ 181
Eclipse IDE as an MCP Server for AI Agents
VSCompare→
turbomcp logo
turbomcp★ 99
A full featured, enterprise grade rust MCP SDK
VSCompare→
mcp logo
mcp★ 78
CLI that turns MCP servers into terminal commands, single binary
VSCompare→
mcp-wayback-machine logo
mcp-wayback-machine★ 66
MCP server and CLI tool for interacting with the Internet Archive's Wayback Machine
VSCompare→
redmine-cli logo
redmine-cli★ 63
A powerful command-line interface for Redmine, enhanced with first-class AI agent capabilities via Skills and MCP integration.
VSCompare→
awesome-mcp-hardware logo
awesome-mcp-hardware★ 60
Awesome list of MCP servers for interacting with hardware and the physical world.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
evalstate /
fast-agent
↗

CLI-first Python framework for MCP-native agents and workflows

4.0k469Apache-2.0
Last commitToday
StatusActive
CategoryDev Tooling
Contributors89
Latest releasev0.10.53
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.