Create or migrate verifiers environments for the Prime Lab ecosystem. Use when asked to build a new environment from scratch, port an eval or benchmark from papers or other libraries, start from an environment on the Hub, or convert existing tasks into a package that exposes load_environment and installs cleanly with prime env install.
Build production-quality verifiers environments that work immediately in the Prime ecosystem: install, load, evaluate, and train without hidden setup.
prime env init my-env
prime env install my-env
prime eval run my-env -m openai/gpt-4.1-mini -n 5
prime eval run as the canonical eval path. It saves results automatically, so do not add --skip-upload unless the user explicitly requests that deviation.prime env list --search "keyword"
prime env info owner/name
prime env install owner/name
prime env install math-python --from-repo
configs/endpoints.toml so smoke tests can switch models quickly.gpt-4.1 series, qwen3 instruct series.gpt-5 series, qwen3 thinking series, glm series.SingleTurnEnv for one-response tasks.MultiTurnEnv for custom interaction loops.ToolEnv or MCPEnv for stateless tools.StatefulToolEnv for per-rollout resources.CliAgentEnv for running agent binaries in sandboxes with API interception. Override get_sandbox_resources(state) for per-instance resources, build_env_vars(state) for custom env vars.ComposableEnv (with TaskSet/SandboxTaskSet + Harness) for separating what to solve from how to solve it. Define a TaskSet (dataset, instructions, sandbox spec, rubric) and a Harness (install script, run command, system prompt), wire them together with zero subclassing. Use SandboxTaskSet when tasks need sandboxes with per-instance images/resources.load_environment(...) -> vf.Environment with explicit arguments.pyproject.toml defaults in [tool.verifiers.eval] only when stable.prime env install owner/name
prime env pull owner/name -t ./tmp-env
load_environment().load_environment() via vf.ensure_keys(...).Run these before claiming completion:
prime env install my-env
prime eval run my-env -m openai/gpt-4.1-mini -n 5
prime eval run my-env -m openai/gpt-4.1-mini -n 50 -r 1 -s
If multi-turn or tool-heavy, also run with higher rollouts:
prime eval run my-env -m openai/gpt-4.1-mini -n 30 -r 3 -s
PUBLIC or PRIVATE.prime env push my-env --visibility PUBLIC
or
prime env push my-env --visibility PRIVATE
prime eval run owner/my-env -m openai/gpt-4.1-mini -n 200 -r 3 -s
Report: