vercel-labs/fx

Unix like coding agent

View on GitHub ↗Jump to charts ↓Open shareable report →

Data as of . Signed-in members get hourly updates — create a free account.

Summary Information

Updated 2 hours ago
Added to GitGenius on September 22nd, 2026
Created on August 11th, 2026
Open Issues & Pull Requests: 239 (+1)
GitHub issues: Enabled
Number of forks: 356
Total Stargazers: 3,142 (+1)
Total Subscribers: 6 (+0)

Repository Insights (GitGenius)

Median issue/PR response: 14.0 hours
Mean response time: 3.4 days
90th percentile: 10.3 days
Tracked items: 75

How this project is maintained

Roughly one issue in three opened in the past year never receives a reply. 100% of open issues come from outside the core team, so the backlog reflects real-world use rather than internal planning. Almost all tracked open issues have seen activity in the last three months. Only 38% of issues opened in the past year have been closed. Three people close 72% of everything that gets resolved.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Open issues: 77
New in 7 days: 13
Closed in 7 days: 10
Avg open age: 21 days
Stale 30+ days: 26
Stale 90+ days: 0

Recent activity

Opened in 7 days: 12
Closed in 7 days: 8
Comments in 7 days: 2
Events in 7 days: 4

Top labels

  • type: bug (25)
  • type: feature (19)
  • bug (17)
  • type: security (6)
  • question (4)
  • type: improvement (4)
  • duplicate (2)

Most active issues this week

Sign in to see which issues are moving.
Sign in

Detailed Description

fx is a coding agent CLI written in Zig that operates as a Unix-like shell interface rather than a terminal IDE.

The tool addresses the need for AI-assisted coding by providing a lightweight native binary that works with any language model. It connects to multiple model sources including the Vercel AI Gateway, ChatGPT, Grok, or self-hosted endpoints like Ollama and OpenRouter. The agent maintains a Unix philosophy by preserving terminal scrollback with inline rendering instead of replacing screen content, and it can be invoked interactively through a shell, as a one-shot command for scripting, or embedded in larger systems through libfx and the ACP protocol.

Developers should choose this tool if they want a model-agnostic agent that integrates naturally into shell workflows without requiring a dedicated terminal UI. It suits projects where you need flexible model selection, including the ability to route requests through different providers via the Vercel AI Gateway or connect to local models. The tool is extensible through skills, MCP servers, and subagents, allowing customization beyond its base capabilities. It works well for teams already comfortable with command-line interfaces and those who want to embed agent functionality into larger systems.

The project maintains active development with regular releases and comprehensive documentation covering sessions, permissions, configuration, and embedding scenarios. The codebase is written in Zig, a compiled systems language, which keeps the binary small and self-contained. The tool is open source under the Apache-2.0 license and explicitly designed to remain model-agnostic, avoiding lock-in to any particular AI provider.