v0.4.0 overview

Agent QC v0.4.0 expands the standard from project case studies into a fuller cross-Agent quality system. It adds Agent UI-inspired runtime-backed surface rules, Agent Skills-style authoring practices, explicit evidence contracts, performance/reliability metrics, and a complete flow/taxonomy reference.

Highlights

• The standard is framed as a protocol rather than a single-product checklist.
• Agent UI surfaces are mapped into QC assertions for composer, message parts, runtime status, tool UI, HITL, task capsules, artifacts, timeline/evidence, sessions, and team workbench.
• Quickstart, gate matrix, classification, acceptance scenarios, and verdict guidance now cover CLI, TUI, WebUI, desktop GUI, browser automation, channel UI, scheduler, release, and eval suites.
• Evidence and waiver semantics are now documented as portable contracts.
• Performance and reliability metrics are documented as QC evidence for streams, tools, browsers, schedulers, and release artifacts.