Repository navigation
learn(demo): live-demo-autopilot — run the demo live with a take-over control panel - #602
Merged
Merged
Conversation
…e-over control panel Merged from two field builds (desktop+phone offline demo 4/4 scenes; three-app hub demo 28/28 with an app killed mid-run and an 11 min pause). Skill text only, on-demand routing row, cross-link from record-demo-video. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018LeNWZV2sVH62vkgfYh5u5
…y default - Logout: a 200 proves nothing. Prove each logout by the System.Session count, and point at testing-shape.md, which measured /xas/ logout returning 200 with the session alive on Mendix 11.14. Keep the field run's timing recipe and fall back to mx.logout() if the count does not drop. - Trial licence stop: 2-4 h, with the 2.5 h run named as one field observation; load-test.md section 9 covers both limits. - Demo sign-in tiles: build them only into a demo or dev build, behind a constant that is off by default, and check it before any deploy. - Live change request: run it on a copy or branch of the demo model. render-routing --check in sync (baseline unchanged, 64,252); check-no-private-citations clean. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018LeNWZV2sVH62vkgfYh5u5
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Direct lane
What changed and why (one paragraph):
New on-demand skill
skills/live-demo-autopilot.md: how to run a customer demo LIVE instead of as a video. Playwright drives the real app through scripted scenes, captions carry a WHY line, and a control panel lets the presenter pause and take over, resume, redo, skip, restart or stop. It merges two field builds: one contributed the desktop+phone layout, offline scenes, per-step traceability (process id + model element) and password-free tile sign-in; the other contributed the reliability layer (gate-based control, idle re-login, an app-health gate per scene, skip recovery, honest X/N summary) and the Mendix traps that each broke a run. The traps: the trial licence stops the runtime ~2.5 h after start; the 6-session cap, and when/xas/logout really frees a slot; a focus-stealing panel pausing the run; SIGTERM closing the browser before sign-out. Skill text only, no code: both engines are macOS-bound, so shipping one would fail the portability bar. Also adds a routing row and a one-line cross-link fromrecord-demo-video.md.Impact
bin/lib/skill-routing.tsv(one new row, tierondemand, groupverify, roletest, stage 6), plus the files it re-renders (README.md,ROUTING.md,agents/test-agent.md,bin/gate-check.sh). No new tools; no artifact readsbin/render-routing.sh --checkoutput is byte-identical on base (origin/master 2ba8158) and head (diffempty), e.g. stage 560203 words / 23 filesand stage 649690 words / 17 fileson both. The row is on-demand, so nothing always-loaded changedField evidence
Desktop + phone offline web-app demo (manufacturing app, 2026-10-20): 4 scenes across 4 users, all scenes passing (commit "Live e2e demo: all 4 scenes pass").
Three-app process-hub demo (2026-10-30): final unattended run
steps passed 28/28 failed 0 gaps 0, with one app killed after scene 3 (the demo logged APP DOWN, waited, and carried on) and an 11 min pause after scene 4. The Stop test took the session count from 3 to 0. The licence root cause comes from the runtime log: start 11:18:58,LicenseService: Maximum run time exceededat 13:48:58.No client data anywhere in the diff (
bin/check-no-client-data.shwith the denylist: no client data detected; the only new proper noun is the contributor credit in CHANGELOG)Size cap: 8 files, 117 insertions / 1 deletion
Test tier: T0 (text and routing).
render-routing.sh --checkexit 0,check-scripts.sh126/126 shell + 27/27 Node parseFor a new/changed instrument: n/a, no instrument
For a new skill: routing row added and rendered
CHANGELOG.mdline appended under Unreleased, crediting the sourceNew bug entries: none
🤖 Generated with Claude Code