read.
This guide uses the filesystem provider because it works anywhere and shows the core Bridge workflow without any extra setup.
Before you begin
- macOS, Linux, or Windows
- A terminal
- A project folder with files you want an agent to read
Step 1: Install Bridge
Install Bridge using one of the supported methods in Installation. After installation, confirm the CLI is available:Step 2: Initialize a Bridge project
Runbridge init in the root of the project where you want to use Bridge.
bridge.yaml file in the current directory. For the file format, URI rules, and manual editing guidance, see Configuration.
Step 3: Connect your first provider
Connect a local directory and name itfiles:
docs folder, point Bridge at any directory you want an agent to read.
Bridge writes that provider into bridge.yaml. For manual editing rules and provider-specific URI examples, see Configuration.
Step 4: List and read context
Usebridge ls to inspect what is available:
What comes next
From here, you can either go deeper into the same workflow or connect another backend:- use Commands to learn the full CLI surface
- use Configuration to understand
bridge.yaml - use Providers to connect Postgres and learn backend-specific behavior
What you have now
- A local
bridge.yamlfile - A named provider connection
- A repeatable way to list and read context
- JSON output that agents can consume directly
Next steps
Commands
See every CLI command, global flag, and current runtime caveat in one place.
Providers
Learn how filesystem and Postgres behave once you move past the first local example.
Back to introduction
Review the core ideas behind Bridge and how it fits into an agent workflow.
View the source
Explore the CLI, roadmap, and release history on GitHub.