Keep every agent you use aligned from one folder in the repo.
# Install agent-layer
$ brew install conn-castle/tap/agent-layer
# Initialize agent-layer in your repo
$ al init
# Launch your agent (sync happens automatically)
$ al claude
Featured clients
Why Agent Layer?
Edit setup once in .agent-layer/. Agent Layer writes each client's config when you launch.
Single Source of Truth
Edit .agent-layer/ once. Skills align with the agentskills.io specification, and the Best Practices guides help keep skills and instructions maintainable.
Sync on Run
By default, al <client> regenerates client configs first, then launches the client. VS Code can skip sync with --no-sync.
Consistent Approvals
Define safety rules once. Agent Layer projects them into each client that supports approvals.
MCP Everywhere
Configure MCP servers in .agent-layer/config.toml. They appear in each client that supports them.
Version Pinning
Pin the repo to a specific Agent Layer version so every machine runs the same release.
Local & Fast
No telemetry. Network is used only for update checks, version downloads (pinning), and MCP servers you enable. See Security for details.
Install Agent Layer
Install on macOS or Linux, then run al init in a repo.
If Agent Layer is useful, consider starring the repository. Stars help new users find the project.