Muse Code journals sessions to an append-only event log. Author a focused golden fixture (records + expected_model_context), then `muse trace inspect --fixture` replays projection with no model, tools, or network. Gate merges on zero diffs via replay_gate.sh / GitHub Actions—catch harness drift before it ships.
Confidee/flowreel (Apache-2.0 Claude Code plugin) turns deterministic Playwright flows into captioned, narrated demo videos with a dev-then-staging release gate—assertions decide pass/fail; Claude writes/explains flows but does not judge them.
Functionize ships an MCP server so Claude Code/Cursor/Copilot hand change context to Studio, which builds/runs smoke or regression suites against the live app and returns the first failing step in-chat—closing the gap where agents never saw the product run.