executing-plans

Executing an implementation plan with dependency analysis involves loading the plan, analyzing step dependencies, and executing tasks in parallel waves. This approach ensures deterministic delivery with automatic gating between waves and per-step verification across code, tests, and documentation.

Parallel wave execution is a plan delivery method that groups independent implementation steps to run concurrently while gating dependent steps sequentially. It enables safe parallel execution by analyzing dependencies first, ensuring steps without conflicts run together before moving to the next wave.

Automated plan execution integrates test-driven development by applying TDD verification to code steps within each parallel wave. The workflow evaluates tests alongside code changes and performs per-step verification before advancing, ensuring implementation aligns with test requirements throughout execution.

Automated plan execution supports frontend and UI design verification by applying UI-change awareness to code steps. The workflow performs design checks during per-step evaluations, ensuring frontend modifications meet visual and functional requirements before the step is marked complete.

Generating execution reports after completing a plan involves running post-run verification and saving a comprehensive execution report to the docs/superomni/executions directory. The report includes per-step results, wave execution status, and automated execution artifacts from the workflow.

Parallel wave execution requires clear dependency mapping to function safely, so plans with ambiguous or circular dependencies between steps may not be suitable. If steps have hidden dependencies or require manual intervention, the automatic gating and per-step verification may halt execution unexpectedly.