sesame request instead of curl or fetch for API calls. Once installed, your agent learns to check which hostnames are configured, handle the first-time approval flow, and never construct auth headers itself — credentials stay inside the broker at all times.
Available Skills
sesame
Teaches your agent to proxy all authenticated API calls through the Sesame broker. The agent runs
sesame request instead of constructing raw HTTP calls, checks configured hostnames via sesame hostnames, and defers all auth header construction to the broker.sesame-onboard
Walks you through deploying a fully self-hosted Sesame broker inside your own AWS account. Covers AWS CLI login verification, Google OAuth setup, a dry-run preview of resources, and the final deploy — all interactively from inside your agent.
Supported Agents
Skills auto-detect your agent platform at install time — no manual configuration required. Supported platforms include:- Claude Code
- Codex
- Cursor
- OpenClaw
- 40+ additional agent platforms detected automatically
Prerequisites
Before installing any Sesame skill, make sure you have:
- The sesame CLI installed on your machine (installation guide)
- Your device registered by running
sesame login - At least one hostname configured in your dashboard at getsesame.dev
Next Steps
- Install skills for your agent — run the installer and get your agent routing through the broker in minutes
- Deploy a self-hosted broker — keep all AI agent traffic inside your own AWS infrastructure