AI Passport — Demo GIF

The “wow moment”: Cursor reads your passport and answers from your real stack — without you re-explaining.

Current file: assets/demo.gif (~21s) — question, MCP get_passport_context, and full AI answer.

Before recording

Record one continuous take. Do not include terminal onboarding — passport is already set up.

Time On screen Action
0:00–0:03 Cursor chat (empty) Optional: flash Settings → MCP → ai-passport ON
0:03–0:08 Chat input Type: What languages and frameworks do I prefer? → Enter
0:08–0:22 Chat response Wait for full AI answer (TypeScript, frameworks…)
0:22–0:25 Same Hold 2s on answer so viewers can read

Pass: AI mentions your passport / coding profile — not generic advice.

Fail: Only the question visible, no answer → re-record.

Quick version (15 seconds)

If you want a short GIF:

Time On screen
0:00–0:05 Question already typed → press Enter
0:05–0:14 Full AI answer visible

Minimum 10 seconds of AI response on screen.

Export settings (ezgif)

  1. Upload OBS MP4 (not URL)
  2. Trim start/end in ezgif Cut video
  3. Convert to GIF:
    • Width: 900px
    • FPS: 12
    • Max colors: 128 (smaller file)
  4. Target file size: < 2 MB if possible, max 5 MB
  5. Save as: docs/assets/demo.gif (overwrite)

Replace in repo

# After saving new GIF:
cd c:\Ai_Passport
git add docs/assets/demo.gif
git commit -m "docs: update demo GIF with AI response"
git push origin main

README and GitHub Pages pick up the new file automatically.

Optional: two-question take (~30s)

  1. What languages and frameworks do I prefer? → wait for answer (~12s)
  2. What project am I working on? → wait for answer (~12s)

Trim to best 25–30s in ezgif if too long.

Troubleshooting

See CURSOR_SETUP.md.

If AI does not use passport: new chat, confirm MCP tools, ask “Use get_passport_context and tell me my coding profile.”