[lagarcia38]

Addresses Jelena's feedback on /sphere/p1-hello-world/.

What changed

• Scenario stated plainly. With Start on and the level reading inside the controller's 300–500 mm band, PLC 1 opens the tank's fill valve; above 500 mm it closes it. The source is the Oct 10 evidence note in sphere-usecases (rockwell-p1-native-integrated-fill-2026-10-10.md).
• Ports labeled.
  • Tank level goes PLC 2 analog output slot 7, ch 0 → PLC 1 analog input slot 7, ch 0.
  • The fill-valve command goes PLC 1 digital output slot 5, point 0 → PLC 2 digital input slot 5, point 0.
• Previous/Next drives every panel. The middle and right values only change at samples 16 and 27 in the data, so stepping looked broken. Each panel now:
  • shows the current sample number;
  • highlights a value that just changed;
  • counts down to its own next change.
• New step 4: what an attacker could do. A four-stage illustration: normal fill, normal stop at 500 mm, a forged 315 mm reading, then overflow. It comes with two check questions. It is described once as an illustration built on the test setup.
• Tutorial shape in the style of the SmartTV guide: an "On this page" list, numbered steps with Expected: results, and the recording's limits stated once in "About this recording". The repeated "did not / not" disclaimers and all-caps labels are gone (about 25 negations → 3 in the visible text).
• Matching edits to docs/p1-rockwell-hello-world.md and the page description.

Checks

• npm run lint: clean.
• npm run lint:links: 0 failures.
• npm run build: OK.
• In-browser on the static export:
  • stepping samples 10→11 and 15→16 updates all three panels;
  • the attack stages render;
  • at 375 px there is no horizontal scroll;
  • the console is clean.

Draft: merging deploys to iotrustlab.com.

🤖 Generated with Claude Code