team-verification-before-completion

What problem does it solve?

Cross-cutting verification guard that enforces evidence-based completion by requiring fresh verification output before any task can be declared done. Applies across all workflow stages: /team-build, /team-qa, /team-ship, /team-debug, /team-tdd, and direct implementation. It covers software tests, ML metrics, pipeline runs, query results, dashboard rendering, and schema validation. Use when an agent claims work is done or signals task completion. Do not use for intermediate progress updates or partial status reports.

Core Features & Use Cases

• Enforces a formal completion gate that requires explicit verification evidence before marking work complete.
• Applies across software development, data pipelines, ML evaluation, dashboards, and analytics for consistent quality control.
• Provides a concrete five-step workflow (IDENTIFY, RUN, READ, VERIFY, CLAIM) to prevent premature completion claims.

Quick Start

Before reporting task completion, trigger the verification gate and execute IDENTIFY → RUN → READ → VERIFY → CLAIM with attached evidence.