Features
Autonomous code experimentation with time budgets
Metric-driven keep/revert decisions
Sandboxed execution with runtime isolation
Git-based artifact replay and reproducibility
Multi-objective optimization (metric vectors)
Pre-programmable experiment specification via program.md
Automated baseline comparison
Verdict generation (VERIFIED·REPRODUCIBLE)
Resilient agent architecture with case-based recovery
Crash-safe commit protocol and deterministic replay
Support for Haiku, Sonnet, and Opus model tiers
Rolling credit billing by wall-clock minutes
Full CLI with authentication and session management
Integration with Python projects (train.py, eval harness)
Local execution (no cloud dependency)
Autonomous vulnerability scanning and PR generation
PR review for quality, security, and best practices
Legacy COBOL to Java migration with testing
Incident triage and root cause analysis
Agent Canvas: manage multiple agents in one workspace
Cloud agents run even when your machine is off
Schedule recurring tasks via cron or event triggers
Trigger workflows from GitHub, Slack, PagerDuty
Model-agnostic: supports Claude Code, Codex, Gemini CLI, MiniMax M2.7
Isolated sandbox execution on VM or cloud
Fine-grained access control and budget guardrails
Audit trails for enterprise compliance
Self-hosted deployment in your VPC
Open-source SDK for custom agent development
In-chat model switching with saved LLM profiles
