#5656th of 118 in Agent frameworks and SDKs
flue
TypeScript framework for building autonomous agents with sandboxes, skills and tools
- Stars
- 8.6k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Last release
- Sep 2026
- Language
- TypeScript
155 on 21 Apr 20268,647 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Flue is a TypeScript agent harness where an agent is a function that composes a model, a sandbox, skills and tools through hooks such as useModel and useSandbox. It provides sessions, subagents, MCP tool connections, durable recovery of accepted work and channel events from Slack, Teams, Discord and GitHub. Agents run locally via CLI or deploy to Node.js, Cloudflare Workers, GitHub Actions, GitLab CI/CD, Daytona or Render.
Who it is for: TypeScript developers building autonomous, sandboxed agents
Strengths
- Agent defined as a single function composing model, sandbox, skills and tools
- Deploys to Node.js, Cloudflare Workers, GitHub Actions, GitLab CI/CD, Daytona and Render
- Durable recovery of accepted work after failures and restarts
- OpenTelemetry, Braintrust and Sentry observability; optional Postgres persistence adapter
Weaknesses
- TypeScript only; no other language runtimes mentioned
- No Docker image or compose file in the repository
- Example uses a hosted Anthropic model; local model support is not shown
- Split across several packages (runtime, vite, cli, sdk, postgres), adding setup surface
What it needs
- no GPU
- Needs Node.js, Postgres (optional persistence adapter)
- Models: Anthropic (claude-sonnet-4-6 in example)
Quick answers
Is flue open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does flue need a GPU?
No, it runs without one.
Also in Agent frameworks and SDKs
See all 118| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 95 out of 100 |
| 2 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 91 out of 100 |
| 3 | mastraTypeScript framework for building AI agents, workflows and MCP servers | 91 out of 100 |
| #4 | browser-usePython library that lets an LLM agent control a web browser | 90 out of 100 |
| #5 | copilotkitSDK for adding agent chat, generative UI and shared state to apps | 90 out of 100 |