Autonomous Loop Engineering: OMNI v0.6.0

We are thrilled to announce the release of OMNI v0.6.0! This milestone release marks a massive step forward in autonomous agent orchestration and production stability, bringing multi-agent loops and zero-allocation performance to your terminal.

Autonomous Loop Engineering

OMNI now provides native support for iterative, autonomous agent loops using OMNI_LOOP_BUDGET and OMNI_LOOP_GOAL environment variables. This enables predictive goal-driven constraints that automatically scale distillation based on your task, so a debug loop keeps its errors while a refactor loop compresses traces.

Maker-Checker Verification Pattern

Scale your tasks cleanly by separating execution from validation. We have introduced the omni_verify MCP tool to securely facilitate the Maker-Checker verification pattern across distinct agent sessions, drastically improving multi-agent task reliability.

Zero-Allocation ANSI & Production Hardening

Performance and stability remain at the core of OMNI:

Consolidated Documentation & Testing

We have unified scattered templates into a cohesive docs/LOOP_ENGINEERING.md master guide, detailing the new multi-agent orchestration capabilities. Furthermore, our test suite has been modernized, renaming all sprint-based tests and achieving 100% coverage across 941 tests!

Update to OMNI v0.6.0 by running omni update.