Installing a bundle creates a private copy you can edit. The published bundle stays unchanged.
109 bundles
Generate realistic variations of workflow data (user inputs, edge cases, adversarial inputs) without real PII
Convert natural language SOPs and runbooks into structured, machine-executable task specifications
Score and bucket tasks by difficulty using baseline agent performance
Extract rare but high-impact failure modes from production logs to create targeted task sets
Order tasks by difficulty, introduce new complexity dimensions progressively
Programmatically produce 10Kâ100K+ task instances from templates, SOPs, and historical logs
Define, constrain, and document the valid action space an agent can take within an environment
Handle environments with persistent state across episodes (databases, file systems, user sessions)
Compose environments spanning multiple tools (IDE + terminal + browser + DB) into a single coherent action space
Create terminal/shell environments with filesystem state, command history, and outcome verification
Set up repo-level coding environments with test harnesses, linting, compilation feedback loops
Construct browser-based sandboxed environments where agents interact with realistic UI surfaces (forms, dashboards, multi-step wizards)