FORG Documentation
FORG gives engineering teams clear visibility and policy controls over AI tool usage. Install in minutes, then use metadata-first dashboards to understand spend, sessions, budgets, and compliance posture.
Up and running in 5 minutes. Install the CLI, activate your license, wire an adapter.
Full REST API docs. Sessions, events, rules, reports, webhooks.
Connect official adapters, planned integrations, and any custom tool through the open protocol.
Understand signals, sessions, the rules engine, and how FORG enforces budgets.
What is FORG?
FORG is a lightweight observability and policy layer for AI coding assistants, CLIs, IDEs, autonomous agents, and internal AI workflows. Official adapters cover core developer tools, while the open adapter protocol lets any HTTP-capable tool emit structured telemetry: model used, tokens consumed, latency, cost, session context, and error codes. No prompts or completions are ever captured; FORG is metadata-only by design.
On top of that telemetry stream, the FORG Rules Engine evaluates policies you define: monthly budget caps per team, per-user token limits, model allowlists for compliance, and real-time anomaly detection. When a rule fires, FORG can block the request, log it, send a webhook, or notify via Slack — configurable per rule. The result is full visibility into your AI spend and the guardrails to keep it predictable.
Get started in 3 steps
One command installs the binary on macOS, Linux, or Windows and removes any old version.
curl -fsSL https://forg.pro/install | sh
Paste the license key from forg.pro/dashboard. The agent stores it in your OS keychain.
forg activate lic_<your-license-key>
FORG detects supported AI workspaces and starts reporting metadata as sessions run. Use the dashboard to confirm live activity.
forg status