Video Walkthrough¶
This 45-second walkthrough shows the real documentation surface, the supported agent setup paths, and the three verified Sentinel DV workflows. It has no narration or audio.
Scene index¶
| Time | Scene | What is shown |
|---|---|---|
| 0:00 | Product boundary | 28 read-only tools, three workflow skills, and no simulator control |
| 0:07 | Agent setup | MCP and project-skill setup for Codex, Claude Code, and GitHub Copilot |
| 0:16 | Regression triage | A real health response with unavailable cohort data excluded from scoring |
| 0:26 | Debug and closure | Bounded waveform evidence and an engineer-reviewed coverage candidate |
| 0:36 | Verification | Deterministic checks for all MCP tools and all three skill workflows |
Reproduce the checks¶
From a development checkout:
.venv/bin/python scripts/verify_all_mcp_tools.py
.venv/bin/python scripts/verify_skill_workflows.py
.venv/bin/python scripts/sync_agent_skills.py --check
The video is a product walkthrough, not a sign-off claim. Sentinel DV reads exported artifacts and returns bounded evidence. It does not run simulations, modify verification sources, or certify coverage closure.