what-we-have-to-work

To bundle unfinished trace.md scenarios, the Skill scans repository docs, collects incomplete items from Implementation Status tables, estimates sizes using a rubric, and groups them into up to three candidate work chunks for selection.

The task planning sizing rubric scales from tiny to xlarge. The Skill reads Implementation Status tables in trace.md files to estimate task sizes and strictly avoids generating bundles that exceed the xlarge threshold.

You can group unfinished development tasks by feature, dependency chain, or shared code areas. The Skill analyzes repository trace files to cluster related scenarios and provides a concise rationale for each grouped work bundle.

Yes, you can use stv:do-work to execute selected work bundles. After scanning trace.md files and presenting up to three candidates, the Skill invokes stv:do-work or stv:plan-new-task based on your choice and backlog viability.

The best way to triage half-complete features is scanning docs/*/trace.md files to collect unfinished scenarios, grouping them into one to three prioritized bundles with size estimates, and presenting them for rapid team selection.

Work bundle estimation might exceed the xlarge limit if too many unfinished scenarios are grouped together. The Skill respects the sizing rubric from tiny to xlarge and avoids exceeding xlarge by splitting tasks into smaller candidate chunks.