What problem does it solve? When running multiple Claude Code sessions in parallel, it is easy to lose track of which session is doing what, apply feedback to the wrong session, or get vague progress updates like "almost done". This Skill enforces a consistent protocol of session headers and explicit status lines so every session is identifiable and every step ends with a clear state. ## Core Features & Use Cases - Session Identification: Prepends a one-line header (project, branch, task) to decisions and plans, with detection of parallel sessions via process checks, git worktrees, and session markers. - Structured Status Lines: Ends major steps with one of four statuses — COMPLETE, COMPLETE_WITH_NOTES, BLOCKED, or NEEDS_INFO — each with a defined format, and NEEDS_INFO always includes a sensible default. - Sprint Dashboard: Compiles a cross-session status view on request, and defines coordination patterns for parallel feature work, independent features, and stacked dependent changes. - Use Case: You run three Claude Code sessions on different branches. Each session reports its branch and task up front, flags a blocked session waiting on credentials, and you get a single sprint view showing which work is ready to merge. ## Quick Start Ask Claude to adopt the sprint-status protocol so every response starts with a session header and ends major steps with a STATUS line.
