← Discover MCPs and Agents
w
MCPAI & MLGitHub

weft-cli

Weft is a developer-first CLI that uses specialized AI agents to build software with structure, discipline, and quality.

Links

Repository

README

From the repo.

Weft CLI

Structured AI workflows for real-world software development

Weft is a developer-first CLI that uses specialized AI agents to help you design, implement, and review software features — while keeping humans firmly in control.

Instead of ad-hoc prompts or “vibe coding”, Weft provides a repeatable, auditable workflow that fits naturally into existing development processes.

What Weft is (and isn’t)

Weft is:

  • A CLI for structured, AI-assisted feature development
  • Built around explicit steps and human review
  • Designed for real projects and real teams
  • Auditable by default

Weft is not:

  • An auto-merge coding bot
  • A chat interface for prompts
  • A replacement for code review or CI

How it works (high level)

Feature request
→ Agents
→ Human review
→ Merge

Each agent has a single responsibility (design, architecture, implementation, validation).
All outputs are written to disk and reviewed before anything is merged.

Quick start

Install Weft:

brew install weft

Initialize it in your project:

weft init
weft up

Create and run a feature:

weft feature create user-auth
weft feature start user-auth
weft feature review user-auth

For full installation instructions and alternatives, see the docs.

Documentation

Start here:

  • Installation → docs/installation.md
  • Agents → docs/agents.md
  • Configuration → docs/configuration.md
  • CLI reference → docs/cli-reference.md
  • Architecture → docs/architecture.md
  • Troubleshooting → docs/troubleshooting.md

Project status

Weft is in early development (v0.x).

The core workflow is functional, but APIs and behavior may change before 1.0.
Feedback and experimentation are welcome.

Contributing

Contributions are welcome.
See docs/development.md for setup and guidelines.

License

MIT License. See LICENSE for details.

Support

Collected info

  • ★ 1 stars
  • Language: Python
  • Source updated: 4/5/2026

Config for your environment

Replace {MCP_ENDPOINT_URL} with this MCP’s endpoint URL (from its repo or docs above). No API key — you connect directly.

Tool

OS

Config file: ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp-server": {
      "url": "{MCP_ENDPOINT_URL}"
    }
  }
}

Paste into mcpServers in the config file. Restart Cursor after saving.

If this MCP is also published on mcpchannel.ai, you can subscribe from Browse and use the gateway config there instead.