Skip to content
RTC Agent LogoRTC Agent LogoRTC Agent

Build from Source

Building from source is ideal for scenarios where you need to develop and debug on the Server side.

Prerequisites

Section titled “Prerequisites”
  • Go 1.27+
  • PostgreSQL 17+ (with pgvector extension) — recommend using the pgvector/pgvector:pg17 image
  • Redis 7+
  • LLM API — Claude or OpenAI-compatible interface

1. Clone the Repository

Section titled “1. Clone the Repository”
终端窗口
git clone https://github.com/rtc-agent/server.git
cd server

2. Start Infrastructure

Section titled “2. Start Infrastructure”

Use the development Docker Compose to start PostgreSQL, Redis, etc.:

终端窗口
# Start dev dependencies (PostgreSQL:15432, Redis:16379, Jaeger, Mock OAuth2, etc.)
go run main.go dev dependencies start

Or start Docker manually:

终端窗口
docker compose -f etc/dev/docker-compose.yml up -d

3. Configuration

Section titled “3. Configuration”

Copy the config template and edit:

终端窗口
cp etc/config.example.yaml etc/config.local.yaml

Edit etc/config.local.yaml, modifying at least the following fields:

database:
  dsn: "postgres://rtc_agent:rtc_agent@localhost:15432/rtc_agent?sslmode=disable"


redis:
  addr: "localhost:16379"


llm:
  provider: "claude"           # or "openai"
  api_key: "${LLM_API_KEY}"   # Injected via env var — set LLM_API_KEY before starting
  model: "claude-sonnet-4-20250514"

Config merging: On startup, the Server automatically loads etc/config.yaml (baseline) + etc/config.local.yaml (overrides). config.local.yaml is already in .gitignore and will not be committed. Only write the diff items.

🔐 API Key security: api_key supports ${ENV_VAR} syntax to reference environment variables, avoiding plaintext keys in config files. Set export LLM_API_KEY="your-key" before starting.

4. Build & Run

Section titled “4. Build & Run”
终端窗口
# Build
go build -o bin/rtc-agent .


# Database migration (first deployment)
./bin/rtc-agent migrate


# Start the service
./bin/rtc-agent serve

The service starts at http://localhost:8888.

5. Verify

Section titled “5. Verify”
终端窗口
curl http://localhost:8888/healthz
# {"status":"ok"}

6. Embed Frontend

Section titled “6. Embed Frontend”

After the Server is running, add the <rtc-agent> component to your web page:

<script type="module" src="https://cdn.example.com/rtc-agent/index.js"></script>
<rtc-agent></rtc-agent>

See Web Component API for details.

Next Steps

Section titled “Next Steps”