Connecting to Claude Code
This is a member-only chapter. Log in with your Signal Over Noise membership email to continue.
Log in to readModule 4 · Section 4 of 8
Connecting to Claude Code
The Kit CLI becomes genuinely powerful when Claude Code agents can call it. The setup is two parts: put the credentials in environment variables, and give your agents permission to run the CLI commands.
In ~/.zshrc (or your shell’s config):
export CONVERTKIT_API_KEY="your-key"
export CONVERTKIT_API_SECRET="your-secret"The kit CLI reads these automatically. Claude Code agents that have Bash tool access can then run kit commands as part of their workflow.
In your Claude Code settings, you’ll need to allow the kit command in trusted commands or run in a context where the agent has shell access. The agent can then do things like:
- Check subscriber growth after sending an issue
- Verify that a tag exists before creating a segment
- Pull broadcast stats for a performance summary
- List current sequences to avoid creating duplicates