progress-log

AI handoff summaries are converted into per-feature progress visibility by propagating new handoff entries into the feature's progress.md and updating its CURRENT header with the latest exit criteria and recommended next steps.

Synchronizing local progress.md with an external tracker uses a local-wins upsert strategy for DAE-managed fields while preserving tracker-managed fields like comments and labels during the TrackedFeature update.

Checkpoint orchestration in agentic workflows recomputes the CURRENT header from the latest handoff's exit_criteria, recommended_next, and any human-blocking unmet criteria to render accurate feature progress state.

Manual reconciliation for an entire project is supported by running a scoped propagation that resolves the methodology root, reads new handoffs, updates progress.md, and performs optional project-wide drift reconciliation.

Tracker synchronization supports mutation testing and ATDD workflows by propagating checkpoint, verification, and fix-closure states from DAE handoff summaries into synchronized TrackedFeature records.

The limitation of local-wins upsert during tracker synchronization is that DAE-managed fields will always overwrite external tracker values, requiring optional drift reconciliation to detect and manage conflicting manual tracker edits.