What problem does it solve? After an implementation pass, teams often cannot tell whether the code actually satisfies everything the spec, plan, and tasks promised. This Skill closes that gap by auditing the current codebase against the feature's spec.md, plan.md, and tasks.md, then appending any unbuilt or incomplete work as new traceable tasks so the implement step can finish it. ## Core Features & Use Cases - Gap Assessment: Classifies every finding as missing, partial, contradicts, or unrequested, with severity levels from CRITICAL to LOW and traceable source references (FR-###, SC-###, US#/AC#, constitution principles). - Append-Only Task Generation: Adds a new numbered Convergence phase to tasks.md with zero-padded task IDs, never rewriting, renumbering, or deleting existing tasks. - Constitution Enforcement: Treats violations of MUST principles in .specify/memory/constitution.md as the highest-severity findings. - Use Case: After running the implement command on a feature, run converge to verify the code covers all acceptance criteria; it either reports a clean converged result or appends the remaining work for the next implement pass. ## Quick Start Ask the agent to run the speckit-converge skill to check the current feature's implementation against its spec, plan, and tasks and append any remaining work to tasks.md.
