AI improvement prompts

Auala e galue ai

E lua pieces of data e tuufaatasi i le click:

• Le finding — title, description, evidence, remediation, CWE — ua uma ona loaded ma le report.
• Le framework o lau codebase — detected mai findings discovery.tech-fingerprint a le scan (Next.js, React, Vue, Django, Express, Rails, Laravel, Flask). Falls back i se generic recipe pe a leai se framework e detected (ona fesili lea le agent prompt i le LLM e detect mai le repo state).

Templates live in lib/scanner/fix-prompts.ts. The registry has check-specific guidance for crawlability, search presentation, semantic content, structured data, media, performance, accessibility, forms, mobile/i18n, runtime, owner journeys, and repo-quality issues. For everything else, the existing remediation field on the finding becomes the generic recipe.

Foliga o le prompt

Fix the "Hero image is lazy-loaded and missing dimensions" finding on /pricing.

Issue: The largest above-the-fold image is marked loading="lazy" and has
no explicit width/height. That can delay LCP and create layout shift.

Codebase context: Next.js.

Recommended fix:
Use next/image or the existing image component with explicit width/height,
responsive sizes, meaningful alt text, and priority/fetchPriority for the
first major visual on the page.

Constraints:
- Don't break existing tests; run the test suite after the change.
- Match the codebase's existing style and lint config.
- Add a brief comment explaining the performance reasoning only where the
  fix would otherwise look arbitrary.
- If the fix needs a new dependency, install it via the project's
  package manager (npm / pnpm / pip / bundle / composer).

Reference: Core Web Vitals / Largest Contentful Paint guidance.

Frameworks e lagolagoina

Matou te surface framework-specific snippets mo:

• Next.js, React, Vue, Nuxt, Svelte (frontend)
• Express, Fastify (Node.js backend)
• Django, Flask (Python)
• Ruby on Rails
• Laravel (PHP)
• ASP.NET Core (planned, fallback to generic today)

Framework detection e best-effort. Matou te sniff tags __NEXT_DATA__, __NUXT__, hash cookies (laravel_session), headers X-Powered-By, ma nisi signals. Afai o e running se custom framework, e fallback le prompt i generic recipe ma iloa e le agent mai lau package.json.

Faaaoga mai lau AI agent

Afai ua e wired up le MCP server, e exposed le same prompt o se slash command. Mai Claude Desktop:

/fixweb-fix finding_id=550e8400-e29b-41d4-a716-446655440000

The renderer looks up the finding, detects the framework from the parent scan when available, renders the templated prompt, and injects it into your conversation as the user message. No round-trip to our Claude API; templates are pure and free.

Aisea matou te le hit Claude ai i click taitasi

I le launch sa matou mafaufau e call le Anthropic API mo click taitasi e refine ai le prompt ma codebase context. Matou te le'i faia, aua:

• O le agent e paste i ai le user ua uma ona iai codebase context — o loo latou faaaoga Cursor / Claude Desktop ma tatala la latou repo.
• Templating per-(check × framework) e cover ~80% o le value e aunoa ma se per-click cost.
• E mafai ona fire le API mulimuli ane e se opt-in “Refine with AI for my codebase” pe a manana'o users. I aso nei, leai.