🧩 fastify Γ— browser-extension

Ship browser extension on Fastify with sensible defaults.

Fastify is a TypeScript workable choice for browser extension. GreatCTO auto-detects both β€” adds the browser-extension archetype overlay, wires browser-extension-specific gates, and runs 83 specialist agents around your existing Fastify workflow.

What changes when GreatCTO joins your Fastify project

Detection β†’ overlay β†’ gates β†’ reviewers.

1 Β· DETECT

Stack + archetype

GreatCTO reads your package.json and detects fastify + browser-extension archetype from signals: imports, file structure, env vars, README hints.

2 Β· OVERLAY

Archetype pack

Attaches the browser-extension archetype overlay: archetype-specific reviewer + compliance gates. Override if your specifics differ; the defaults are sensible for Fastify-style projects.

3 Β· GATES

Fastify-aware reviewers

qa-engineer runs tsc --strict / eslint / vitest --coverage; security-officer checks for prototype pollution + XSS sinks; performance-engineer reviews bundle size + cold-start times.

4 Β· MEMORY

Cross-project lessons

Bugs you've hit before in other Fastify projects (connection-pool exhaustion, ORM N+1 queries, retry storms) β€” the agent's Step 0 includes the prior detection order. MTTR drops 94 % on second occurrence (methodology).

First 10 minutes

Concrete walkthrough.

$ cd my-fastify-app && npx great-cto init
βœ“ scanning manifests… found package.json
βœ“ stack: fastify (TypeScript)
βœ“ archetype: browser-extension
⚠ archetype + stack combo is unusual β€” review overlay manually
βœ“ 83 agents ready

$ /start "add browser-extension feature"
β–Έ architect drafting ARCH-browser-extension.md…
β–Έ pm decomposing into beads tasks…
⚐ gate:plan β€” your approval needed

Approve β†’ 3 senior-devs run in parallel worktrees β†’ 5 reviewers fan out in parallel β†’ gate:ship β†’ deploy. One real run walked stage-by-stage: /proof.

What ships

The first feature, Fastify-idiomatic.

This is the shape of what senior-dev drafts for "browser-extension feature" β€” auth first, schema validation, and the audit line the browser-extension reviewer requires before gate:ship opens.

// src/routes/browser-extension.ts β€” drafted by senior-dev, reviewed by 5 agents
export default async function routes(app: FastifyInstance) {
  app.post('/browser-extension', {
    preHandler: app.requireAuth,              // security-officer: auth before handler
    schema: { body: browserExtensionSchema },  // qa-engineer: JSON-schema enforced
  }, async (req, reply) => {
    const result = await handle(req.body, req.user);
    await app.audit.log(req.user.id, 'browser-extension feature', result.confidence); // gate:browser-extension
    return result;
  });
}
Where this combo lands

What teams build with Fastify + the browser-extension overlay.

1

Productivity extensions with minimal host permissions.

2

Enterprise extensions passing Chrome Web Store review.

3

Cross-browser extensions (Chrome / Firefox / Safari).

⚠ Honest caveat

Fastify (TypeScript) is not a typical fit for browser extension. The archetype overlay still attaches, but you may want to override defaults more aggressively. Check the browser-extension archetype page for the typical stack list and decide if your case is the right tool / right archetype.

Architecture

Every step of the pipeline, transparent.

No black-box "AI does it all" loop. GreatCTO is a deterministic state machine β€” 8 stages, 22 nodes, 2 human gates. Every node maps to a real agent on GitHub. Inspect the state machine β†’

Install

Fastify + GreatCTO in one command.

$ npx great-cto init

Free, MIT, runs locally. Built as a Claude Code plugin β€” install with one command.

Related deep-dives

More from the blog

AI

How I designed the SDLC state machine for agentic coding

Eight stages, two human gates, four memory layers. Why this exact shape, and what I tried that didn't work.

AI

First real shipped feature with this stack β€” receipts

One run, one feature, from prompt to merged PR. Time, cost, and gate-by-gate breakdown β€” no marketing math.

AI

How GreatCTO chooses which compliance pack to attach

Regex vs LLM-based archetype detection, the false-positive count, and why I keep rejecting the obvious fix.

AI

Why your agent system fails: missing gates, not missing intelligence

The bottleneck in agentic SDLC isn't model quality β€” it's process governance. Here's the state machine that closes the gap.