Start integrating in minutes.
REST APIs, Python and TypeScript SDKs, CLI tools, and step-by-step guides. Build evaluation, training, and monitoring into your existing workflow.
One API per product. Full lifecycle access.
Evaluation, training, deployment, and monitoring — all accessible programmatically. RESTful endpoints with OpenAPI specs.
Environments, agents, training runs, certifications, and deployments. Full lifecycle management via REST.
Register agents, start runs, call tools, submit evaluation results. Programmatic access to all benchmarking features.
Live certification results, drift alerts, cost breakdowns, and webhook configuration. Real-time fleet visibility.
Use the SDK that fits your stack.
Python for Forge integration. TypeScript for Agent 007. CLI for terminal-first workflows.
pip install xplore-forge. Environment setup, training runs, evaluation chains, and deployment management.
npm install @xplore/agent-007. Agent registration, benchmark execution, and result submission.
xplore-cli for terminal workflows. Create environments, trigger training, check certifications, promote agents.
Go from first call to production setup.
From first evaluation to production monitoring. Each guide maps to a verb: Evaluate, [Re]train, Deploy, Control.
First evaluation in 5 minutes. Connect an agent, build an eval chain, run a benchmark.
Configure training strategies, promotion policies, and auto-retrain triggers step by step.
Set up live certifications, drift detection, cost guardrails, and alerting for deployed agents.
Event notifications for training completions, certification results, drift alerts, and promotions.
Control who can access what — and see who did.
Generate API keys with granular scopes — read, write, admin. Every API call logged with actor, timestamp, and scope. Rotate keys without downtime.