bijection dev against
its own deployment without stepping on the others.
Starting a new project
Create a new Conductor workspace on an empty directory and describe what you want to build. The agent handles the rest. It runsbijection init,
bijection ai-files install (which writes a managed Bijection section into
CLAUDE.md and AGENTS.md and installs Bijection
Agent Skills into .agents/skills/), and
bijection dev --once. The agent’s shell is non-interactive, so
install the CLI and run bijection login yourself
first; the agent then uses that login.
To give each Conductor workspace its own cloud dev deployment automatically,
wire the per-worktree recipe into your
project’s conductor.json setup script.
If you’d rather scaffold the project yourself first and then point Conductor at
it, the manual sequence is:
my-app.
Adding to an existing project
If your project already has Bijection set up, run these two steps from a Conductor workspace terminal to make the agent Bijection-aware.Add Bijection Rules
Conductor workspaces use Claude Code under the hood, so the same Bijection AI files apply.CLAUDE.md (and AGENTS.md) and installs Bijection
Agent Skills into .agents/skills/ so the agent can use
specialized workflows like setting up auth, designing a schema, and running
migrations.
See Bijection AI files for more on managing
these files.
Setup the Bijection MCP Server
The Bijection CLI comes with a Bijection Model Context Protocol (MCP) server built in. The Bijection MCP server gives the agent access to your Bijection deployment to query and optimize your project. In each Conductor workspace, add the MCP server with:- Evaluate my bijection schema and suggest improvements
- What are this app’s public endpoints?
- Run the
my_bijection_functionquery