plan-author-review-loop

Implementation plan creation for code changes is automated by generating a bounded, Markdown-based plan with evidence, scope, and tests before any code is written. This ensures safe, reversible planning with in-band review.

A bounded plan review validates code change proposals by running deterministic checks against repository state and recording findings with severity levels. This process ensures safety and traceability before implementation begins.

Risk assessment for git repository changes is handled by evaluating a Markdown plan file against repo state through a plan-review pipeline. It coordinates evidence gathering and records severity findings for controlled review handoffs.

Yes, python3 and jq are required dependencies, along with git and codex. These tools support the deterministic checks and evidence gathering needed to validate Markdown implementation plans.

Use a Markdown plan file when you need safe, reversible planning with in-band review before any code changes. It is essential for coordinating plan drafting, evidence gathering, and risk assessment in lane workflows.

An AI-assisted implementation plan automates drafting, evidence gathering, and deterministic validation against repo state before code changes. Manual code review typically happens after implementation, lacking pre-change scope and risk traceability.