Planning

Enhance's planning system helps you understand how issues interact and determines the optimal order to work on them. It combines static file overlap analysis with AI-driven execution planning.

Overlap analysis

When multiple issues affect the same files, working on them simultaneously can cause merge conflicts. Enhance computes file overlap between all pairs of open issues.

Each overlap pair includes:

• Overlap score (0-1) — How severely the issues conflict
• Shared files — Which files both issues will touch, with probability from each issue
• New file predictions — Predicted files that both issues might create

This analysis runs on existing impact analysis data from the issue processing pipeline.

AI execution planning

The AI planner takes your issues, their dependencies, and overlap information, and determines the optimal execution order:

• Batches — Groups of issues that can be worked on in parallel
• Ordering — Sequence of batches (batch 1 completes before batch 2 starts)
• Rationale — Why each batch is grouped together
• Effectiveness score — How well the plan avoids conflicts

Strategy controls

Option	Values	Description
Overlap strategy	Together / Sequential	Whether overlapping issues should be in the same batch or separate ones
Sequential commits	On / Off	Whether to enforce one commit per batch
Max issues per batch	Number or unlimited	Limit parallelism within a batch

Plan editor

The plan editor provides a visual canvas (powered by ReactFlow) where you can:

• See issue cards arranged by batch
• View dependency edges between issues
• Drag issues between batches
• Review the overlap analysis visualization
• Check effectiveness scores and warnings

Integration with agents

Once a plan is finalized, it can be executed by AI development agents. Each issue in the plan is assigned an execution order and tracks dependencies on other issues. The agent system processes issues batch by batch, respecting the planned order.