Fast Agent
An open-source coding agent and Python toolkit that runs in your terminal, with llama.cpp support, cloud model providers, and MCP tools.
Fast Agent is a terminal coding assistant and Python toolkit for developers who want to build and evaluate AI agents with their choice of model. It supports local LLMs through a generic provider and automatically configures connections to llama.cpp servers. The toolkit runs on your machine; choosing Anthropic, Google, or a remote OpenAI-compatible endpoint sends model requests to that service.
You can use it interactively with shell tools or build scripted automations through its Python API. MCP support connects agents to external tools and resources, while ACP and MCP let you expose agents to other clients. Python functions can also act as tools directly, without a separate MCP server.
Its workflow options cover sequential agent chains, parallel work, and routing tasks to an appropriate agent. An orchestrator can divide a larger task among agents and combine their results. For work that needs revision, an evaluator can judge a generator's output and request improvements in a loop. Agents can ask a person for additional context during a task.
Context controls include system prompt templates, Agent Skills, and reusable prompt files. Python plugins and hooks let developers extend behavior, and Card Packs distribute agent configurations. Multimodal prompts and MCP resources support content beyond text, subject to the selected model provider's capabilities. Fast Agent is open source under Apache 2.0.
