The service three teams depend on was documented once, two years ago, by someone who left.
Authentication
Credentials resolved at the runtime, never in agent code
Idempotency
A retry never repeats a real action
Validation
Responses checked against the schema before your agent acts
Audit
A record of what was requested, allowed, and executed
CLI
Pull the manifest, set your policy, execute. No SDK sprawl, no boilerplate. Just the tool you need, right now.
swytchcode get github
swytchcode exec github.create-pr \
--repo org/repo --head feat/ai-agent \
--base main --title "Auto: refactor auth"How it connects
Follow these steps to connect your agent, load a toolkit, and start executing tools through Swytchcode.
Install the CLI and authenticate so your dev tooling can route through a single execution layer.
npm install -g swytchcodeStart the MCP server so your IDE agent can discover the Swytchcode toolkit.
swytchcode mcp serveLoad the integration your agent needs, such as GitHub.
swytchcode get githubLet the agent generate a PR or triage a bug while Swytchcode validates every execution.
swytchcode exec github.create-pr --repo org/repo --head feat/ai-agent --base main --title "Auto: refactor auth"Works with
Install the CLI and pull your first manifest in under a minute. No account required.