Self-hosted ChatGPT-style chat platform that unifies many AI providers, with Agents, MCP support, Artifacts, and multi-user auth.
danielmiessler/Fabric
An open-source Go framework that organizes reusable AI prompts (called Patterns) by task and runs them from the command line or a REST API.
- Stars
- 44,359
- Forks
- 4,308
- Open issues
- 18
- Last commit
- 22 hours ago
What is it?
Fabric is a framework for applying AI to everyday work by turning prompts into reusable units. Each task-specific prompt is stored as a Pattern, such as summarizing an article, extracting wisdom from a YouTube video, or explaining code, so users can collect and reuse their best prompts in one place instead of rewriting them in each chat app.
The command-line tool takes text from stdin or files, applies a chosen Pattern, and sends the result to a model from one of many supported providers, including OpenAI, Anthropic, Google Gemini, Ollama, Azure OpenAI, Amazon Bedrock and OpenRouter. Patterns are plain Markdown files that can be customized, and the tool can also pull YouTube transcripts, scrape web pages, and transcribe audio.
Fabric can also run as a REST API server, and in Ollama-compatible mode it lets Ollama clients use Fabric's providers and Patterns. It is written in Go and distributed under the MIT license.
Key features
- 01Organizes prompts as named Patterns that can be run by task.
- 02Accepts input from stdin, files, YouTube videos, web pages and audio.
- 03Supports many AI providers, including local models through Ollama and LM Studio.
- 04Includes a REST API server and an Ollama-compatible mode.
- 05Offers dry-run and prompt export options for inspecting prompts before sending them.
- 06Provides shell completions for Zsh, Bash and Fish.
Use cases
- Summarizing articles, papers or long videos from the terminal.
- Building a personal library of reusable prompts shared across tools.
- Running prompts against local or cloud models from scripts and pipelines.
- Exposing a single API endpoint that routes requests to several AI providers.
Quick start
- 1
Install Fabric with the one-line installer on Unix, Linux or macOS
curl -fsSL https://raw.githubusercontent.com/danielmiessler/fabric/main/scripts/installer/install.sh | bash - 2
On Windows, install Fabric with PowerShell
iwr -useb https://raw.githubusercontent.com/danielmiessler/fabric/main/scripts/installer/install.ps1 | iex - 3
Run the setup to configure your AI provider keys
Things to know
- You need API keys or accounts with supported AI providers to run patterns.
Topics
FAQ
- What is Fabric?
- An open-source Go framework that organizes reusable AI prompts (called Patterns) by task and runs them from the command line or a REST API. Fabric is a framework for applying AI to everyday work by turning prompts into reusable units. Each task-specific prompt is stored as a Pattern, such as summarizing an article, extracting wisdom from a YouTube video, or explaining code, so users can collect and reuse their best prompts in one place instead of rewriting them in each chat app.
- Is Fabric free to use? What license does it have?
- Fabric is open source under the MIT license. Check the license terms before using it in commercial products.
- Is Fabric still maintained?
- Actively maintained. The last commit to Fabric was on Oct 18, 2026.
- What can I use Fabric for?
- Summarizing articles, papers or long videos from the terminal. Building a personal library of reusable prompts shared across tools. Running prompts against local or cloud models from scripts and pipelines. Exposing a single API endpoint that routes requests to several AI providers.
- How do I install Fabric?
- Install Fabric with the one-line installer on Unix, Linux or macOS: curl -fsSL https://raw.githubusercontent.com/danielmiessler/fabric/main/scripts/installer/install.sh | bash On Windows, install Fabric with PowerShell: iwr -useb https://raw.githubusercontent.com/danielmiessler/fabric/main/scripts/installer/install.ps1 | iex Run the setup to configure your AI provider keys
- Where can I get Fabric?
- The source code and installation instructions are on GitHub: https://github.com/danielmiessler/Fabric
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
An LLM-powered web scraper, built on Streamlit, that extracts data from pages using OpenAI, Gemini, Ollama or LiteLLM models.
A CLI and web dashboard that benchmarks free coding LLM models across providers and installs the chosen one into your coding tool.
A Zotero plugin for chatting with single or multiple PDFs and searching the whole library, using cloud LLMs or local models.
Browser automation platform that turns recorded browser actions into MCP commands or Claude Skills so AI agents can control browsers.
A CLI that scans your codebase and generates compact AI context files (CLAUDE.md, .cursorrules, AGENTS.md) so coding assistants spend fewer tokens exploring the project.