Spec your site with the Gem
Before you design or code, give the AI a clear brief. The workshop Gem interviews you for ~10 minutes and outputs two ready-to-paste prompts — one for Google Stitch (next step) and one for Google Antigravity (step 4). One tool, one conversation, two prompts out.
A Gem is a Gemini chat with a fixed system prompt and attached knowledge files. The workshop Gem has a disciplined one-question-at-a-time interview script and a research file baked in (recruiter expectations for 2026, regional employer references, AI-disclosure norms). You get sharper questions and grounded prompts without having to do the research yourself.
Where the research came from: the instructor pre-built a NotebookLM with ~60 sources on portfolio hiring in 2026, then condensed the relevant findings into a single context file attached to the Gem. If you want to explore that corpus yourself, see the NotebookLM deep dive — it's optional self-study and not part of the live workshop.
Open the Gem
Open the workshop Gem in a new tab and start a fresh chat.
The link looks like gemini.google.com/gem/<workshop-gem-id>. Sign in with the same Google account you used in step 1.
The Gem won't start automatically. Send this message in the chat to kick off the interview:
Let's start interviewing me.Your interview answers will include career goals, target employers, projects, and personal context. How that data is used depends on which Google account you sign in with:
- Personal Google account (free Gemini): by default, your conversations are used to train Google's models and may be reviewed by humans. Default retention is up to 18 months; reviewer-flagged conversations may be retained up to 3 years.
- Workspace / .edu account (paid): conversations are not used for training outside your domain and are not human-reviewed without permission. Treated as customer data under the Workspace DPA.
Opt-out for personal accounts (~30 sec): open myactivity.google.com/product/gemini and turn Keep Activity off before starting the interview. You can also delete the chat afterward.
If you'd rather not run the interview through the Gem at all, use the manual alternative below.
The interview
The Gem walks you through 8 areas in roughly 6–8 questions (one at a time — some get bundled when you give rich answers):
- Identity — name, current city and country
- Target market — the most important question. Which country (or kind of role) do you want to start your career in? Local, EU-remote, US-remote, or hybrid? The Gem classifies you into one of eight Market Profiles (A–H) that drives every later choice.
- Target role — be precise. "Frontend dev" is too vague; "React frontend dev focused on accessibility" lands.
- Audience — who must this site convince? Local recruiters, EU-remote hiring managers, freelance clients?
- Experience — internships, freelance, hackathons, open-source PRs, with concrete numbers where possible
- 3–5 projects — for each: problem, stack, your specific contribution, how you used AI, the result
- Differentiator — what makes you memorable in a sea of "junior full-stack developer" applications?
- Voice + language + visual identity + constraints — tone, languages the site needs (English-only, bilingual EN+FR, EN+TR, etc.), color palette, fonts, light/dark, domain, anything you explicitly don't want
Answer in two or three sentences each. The site is a launchpad, not an autobiography.
"I want a job somewhere in Europe" gives a generic site. "I'm in Cairo and I want EU-remote roles at Series B SaaS scaleups" tells the Gem to put you in Profile B and bake in the right time-zone badge, the right employer references (GitLab, Hotjar, RevenueCat), and the right credential signals (open-source PRs, async-comm artifacts).
Out: two prompts
When the Gem has covered all 8 areas, it summarizes and asks "Ready to generate your Stitch and Antigravity prompts?" Reply yes, and it outputs two fenced code blocks under two short headings:
- Stitch prompt — used in step 3 to generate the visual design
- Antigravity prompt — used in step 4 to generate the actual site code
Each block starts with a comment line like # Market profile: B — EU remote from outside EU so you (and the downstream tools) can see which framing was applied.
Use the Copy button (top-right of each fenced block) to grab one prompt at a time. Save both into a text editor or notes app you can paste from later. Once your project is scaffolded in step 4, save the Antigravity prompt to docs/antigravity-prompt.md at the project root — the agent references it for re-grounding when it drifts.
Alternative — write the prompts yourself in an editor
If you'd prefer not to share the interview content with Google, you can self-author the two prompts. This takes longer (~25 min instead of ~15) but gives you full control.
The Gem's job is to (a) extract the same eight areas covered above, (b) classify you into a Market Profile, and (c) inject profile-specific employer references, time-zone badges, and credential signals into both prompts. To do it yourself, look at the workshop repo for template/guide/gem/workshop-research-context.md — it has the per-profile briefings you'd need to write the prompts grounded the same way the Gem does.
Match the structure: one prompt for Stitch (design — sections, layout principles, palette, employer mentions, anti-patterns), one for Antigravity (build — stack, file scaffold including persona.json / Dockerfile / GEMINI.md / AI-USAGE.md, region-specific add-ons, final commit).
Iterate if needed
If the prompts feel off, re-run the Gem — it's free and fast. The most common cause of a generic prompt is a vague Market Profile answer. Better to fix it here than to discover misalignment in step 4 when Antigravity is two prompts deep.
Why this step exists
Decoupling content from layout. When you spec first, you can iterate on visuals (Stitch) and code (Antigravity) without re-deciding what the site is about. The spec is the source of truth; everything downstream is presentation.
Better AI input through grounded research. A generic "build me a portfolio" prompt produces a generic site. A prompt that says "make this Tunis-based engineer look credible to Berlin Series B fintech recruiters, with a French/English language toggle and a GMT+1 time-zone badge" produces something specific — and that specificity comes from the research baked into the Gem.
- The Gem runs a short interview and outputs both the Stitch and the Antigravity prompts directly.
- The most important question of the interview is your target market — it drives every later personalization.
- Save both prompts before moving on — Stitch in step 3, Antigravity in step 4.
- NotebookLM is no longer in the live workshop path. The research it produced lives inside the Gem; explore the notebook itself in the optional NotebookLM deep dive.