progress-tracking

Checkpoint your progress proactively at key moments—task completion, discoveries, planning decisions, and major changes—to create a persistent memory that survives crashes and long gaps. The skill auto-detects these moments, generates descriptions, tags them, and records metadata like timestamp, branch, and changed files for later recall.

Checkpointing captures specific, action-oriented descriptions of completions and decisions with 2-4 tags, building a clear narrative for standups and recalls. It integrates with plan workflows to tie checkpoints to active roadmaps, ensuring team alignment on progress and decisions.

Yes. The skill operates proactively during feature implementation, testing, bug fixes, refactoring, and before risky changes—automatically detecting checkpoint moments and capturing progress without requiring user permission.

By labeling checkpoints with precise descriptions and tags at the moment of completion, you eliminate the need to reconstruct context later. Checkpoints preserve what you discovered, decided, and completed, making standups and session handoffs straightforward.

The skill generates specific, action-oriented descriptions automatically at checkpoint moments. You can optionally instruct it directly—for example, "Checkpoint: Implement login flow, tags: [feature, auth]"—to integrate checkpointing into your workflow.