Beta · 5 pilot seats

One release gate for everything that breaks a website.

Security audit, performance, release smoke tests and SEO — behind one API key. One report and one pass/fail answer before you ship, instead of four tools and a spreadsheet.

Runs only against sites you own or are authorized to test.

GET /quality/healthzchecking
$ curl https://ai.nocturndev.com/quality/healthz connecting to production gateway…
Free · no signup

Scan any website. Security and SEO in one report.

Security: availability, TLS, security headers and cookies. SEO: a 10-page crawl with on-page, technical, PageSpeed, server, sitemap and backlink checks. Takes about a minute. Only ordinary page requests, the same ones a browser or search crawler makes. Code audit needs repository access and is part of the pilot.

Coverage

Four checks that usually live in four different tools.

Each check runs as its own service. You get one base URL, one key, one contract — and a single answer to “is this release safe to ship?”

Security audit

Rules written from real production incidents, not a generic checklist.

  • Leaked secrets & env exposure
  • Auth and tenant boundaries
  • Dependency & supply chain

Performance & capacity

Controlled, low-limit load on authorized targets with real numbers.

  • p95 latency & error rate
  • Availability under burst
  • Rate-limit boundary checks

Release smoke

The checks you forget at 6 pm on a Friday deploy.

  • Health, readiness, API contract
  • Accessibility & visual diff
  • Critical browser flows

SEO & AI visibility

Technical SEO that silently regresses between deploys.

  • Crawl, index, sitemap, robots
  • Soft 404s & broken metadata
  • llms.txt & AI-search readiness
Today

Four dashboards, zero verdicts

Security scannerseparate login
Lighthouse / PageSpeedmanual runs
Uptime & synthetic checksseparate bill
SEO crawlerweekly export
With Nocturn Quality

One call after every deploy

/v1/auditsecurity
/v1/smokerelease & capacity
/v1/seosearch & AI visibility
/v1/usageone key, one bill
release gate: PASS — safe to ship
Dogfooding

We ran it on ourselves first.

Before offering it to anyone, we pointed the full stack at our own monorepo and APIs.

1,000virtual users in a controlled smoke
0.00%errors under an isolated high-rate profile
4.5 msp95 latency, same run
1false-positive class found in our own tool — and fixed
The first run flagged our own generated reports as application source in a monorepo. We fixed the collector, added a regression test, and re-ran. The default rate limit also correctly rejected an unthrottled burst with HTTP 429 — exactly the kind of boundary a release smoke should expose.

Local, isolated test profile. Not a penetration test, a production capacity guarantee, or an independent security certification.

API

Built for CI, not for screenshots.

1

One key

Bearer token or X-Nocturn-Api-Key. Every request gets an x-request-id you can trace.

2

One contract

Unified OpenAPI at /openapi.json, namespaced services under /v1.

3

Fails closed

/readyz returns 503 the moment any upstream is down, so your pipeline never ships on a half-checked release.

# gate a release in your pipeline
export NQ=https://ai.nocturndev.com/quality

curl -fsS $NQ/readyz

curl -fsS $NQ/v1/smoke/smoke-runs \
  -H "Authorization: Bearer $NOCTURN_KEY" \
  -H "Idempotency-Key: $GIT_SHA" \
  -H "Content-Type: application/json" \
  -d @release-gate.json  # mode + target config

# 202 Accepted → poll the Location header

# full reference → /quality/docs
Plans

Free to try. Cheap to keep.

The keyless scan stays free. When you want an API key, higher volume and the repository security audit, pick a plan — request access and we email your key.

Loading plans…

Beta pilot

14 days. One staging target. A before/after report.

For agencies that send clients regular web reports, small SaaS teams, and e-commerce teams without a pre-release quality check.

first 5 teams
$49 / 14-day pilot
  • Baseline quality snapshot with prioritized risks
  • Security, SEO & AI-visibility findings
  • Release smoke gate wired into your CI
  • Controlled capacity test with p95 and error rate
  • Day-14 comparison: new, fixed and remaining issues
Request a pilot seat

Optional afterwards: $99/month beta price. Custom integration: $99 setup.

Ground rules

  • We only test targets you own or explicitly authorize.
  • Load tests start at low limits; active tests on production need written approval.
  • No credential stuffing, wordlists or attack simulation.
  • Reports are sanitized — no credentials or raw source in the output.
  • In exchange for the beta price: honest feedback. A reference is welcome, never required.