What - a runnable example per feature plus the common end-to-end workflows.
Why - seeing a feature wired up end to end, config and all, is faster than reading a reference and guessing.
How - each directory under examples/ is self-contained with its own README; copy the one closest to your setup and adapt it.
Most examples ship a Docker Compose file:
| Example | Demonstrates |
|---|---|
| basic | Minimal gateway + CLI setup to get started |
| a2a | Agent-to-Agent: multiple agents, a demo site, and a VNC container |
| mcp | MCP server integration with a sample server and config |
| tools | Custom tools in Python and shell, run offline against a scripted mock model |
| computer-use | Computer Use driving a sandboxed Ubuntu GUI container |
| model-switching | Switching models mid-session, with a small frontend |
| shortcuts | Custom /-shortcuts wired through config |
| web-terminal | Browser-based, multi-tab web terminal |
| telegram-channel | Driving the agent from a Telegram channel |
| working-offline | Fully offline usage with local models via Ollama or llama.cpp |
| gpu-provisioning | Renting an on-demand cloud GPU running llama.cpp via infer gpu (RunPod) |
| postgres-storage | Persisting conversations to PostgreSQL |
| a2a-traces | End-to-end OpenTelemetry traces between the CLI and an A2A agent |
| a2a-gateway | Every A2A agent behind the gateway, addressed by tenant, with a mock model and no API key |
| a2a-auth | A2A agents behind a bearer token and behind OIDC with Keycloak, with a mock model and no API key |
| a2a-auth-gcp | An A2A agent behind Google Cloud identity, authenticated with a service account ID token |
| a2a-auth-entraid | An A2A agent behind Microsoft Entra ID, with the OIDC client-credentials grant and az tokens |
| a2a-auth-aws | An A2A agent behind Amazon Cognito, authenticated with a machine-to-machine token |
| a2a-gateway-auth | The gateway applying OIDC auth, guardrails and tracing to A2A traffic, with a mock model and no API key |
There is also an examples/kubernetes manifest set for running the gateway and an agent on a cluster.
# Start chat
infer chat
# In chat, use shortcuts to get context
/scm issue 123
# Discuss with AI, let it use tools to:
# - Read files
# - Search codebase
# - Make changes
# - Run tests
# Ask the agent to open the PR when ready
> Create a pull request for these changes - it fixes the authentication timeout issue- Quick Start - first chat in three commands
- Commands Reference - the main commands and the global flags
- Directory Structure - what the CLI writes where