Vibe 101 Session Flow
Station 2 / 7
BMFeedback

Start Here · 02 / 07

Vibe 101 Session Flow

The 3-hour SDD Website Build Lab map, with timeboxes and the output you finish each stage with.

Vibe 101 / Current checkpoint

Session Flow

Ready to stamp - Saved in this browser only.

0 of 7 checkpoints complete

0 of 7 checkpoints complete.

Vibe 101 Session Flow

What this is

This is the 3-hour map for the SDD Website Build Lab. You use the Vibe SDD Starter Pack in Website mode and move through one saved stage at a time:

Flow
Setup Gate -> Website mode -> Project Brief -> Design Direction -> Build Blueprint -> Work Cards -> First Visible Build -> Review/Fix -> Proof

Why it matters

The session moves fast, so each stage needs a clear output. Saved files protect the plan, and Work Cards keep implementation small enough to check and explain.

Aim for one of these success levels:

Copy
Minimum: SDD planning files + first visible website or screenshot + 60-second explanation.
Good: minimum + one Review Mirror fix.
Strong: a GitHub and Vercel URL when time and accounts allow.

What to do

Follow this timeboxed agenda:

TimeStageWhat you finish with
0:00-0:15Open the Starter PackWebsite mode chosen
0:15-0:30Setup GateTools, accounts, and folder ready
0:30-0:50Choose your website projectOne small, locked idea
0:50-1:15Project Brief via the Blueprint Coachproject-brief.md saved
1:15-1:35Design DirectionDesignMD inspiration translated into design.md
1:35-1:55Design + Build Blueprintbuild-blueprint.md, build-status.md, and first card saved
1:55-2:30First Visible BuildFirst visible website section
2:30-2:45Review + FinishOne checked fix against brief and design
2:45-3:00Capture proofScreenshot or URL plus 60-second story

Keep the whole build inside the shared Vibe SDD files. If a chat or credit limit interrupts you, build-status.md tells the next session exactly where to continue.

Common mistakes

  • Starting the build before the Setup Gate and planning files are complete.
  • Trying to implement several Work Cards at once.
  • Letting AI skip explanations or verification.
  • Treating deployment as the only valid finish.