Skip to main content
One real run · 31 March 2026

We pointed P1·QA at P1·QA.

It scored 61.8%. Grade C. We are publishing the whole thing, because a QA company that only shows you its wins is not a QA company.

Target https://p1qa.com · 3 agents · 36 checks · artifact reports/P1QA-demo-audit-2026-03-31.json

61.8%

Overall pass rate

Grade C

20

Findings on our own site

1 critical · 5 high · 11 medium · 3 low

36

Checks executed

24 passed · 12 failed

66.5s

Wall-clock runtime

3 agents, start to report

Agent by agent

The free audit runs three agents. Our accessibility scored an A. Our SEO scored an F. Both numbers are below, unedited.

AgentDisciplineGradePass rateChecksPassedFailedFindings
AegisSecurityC68.8%1611510
PrismAccessibilityA91.7%121111
IndexerSEOF25.0%8269
OverallC61.8%36241220

The 5 high-severity findings

Verbatim from the 31 March 2026report — and what we shipped since.

All 5 resolved as of 31 August 2026
  • SEC-001Missing content-security-policy headerSecurity HeadersFixed — CSP shipped with a report-uri endpoint
  • SEC-002Missing strict-transport-security headerSecurity HeadersFixed — HSTS, max-age 1 year, includeSubDomains + preload
  • SEC-009No clickjacking protectionClickjackingFixed — X-Frame-Options: DENY + frame-ancestors 'none'
  • SEC-010Security header score 10/100 — Grade FSecurity ScoreFixed — Full header set: CSP, HSTS, XFO, nosniff, Referrer-Policy, Permissions-Policy
  • SEO-001Missing meta descriptionTechnical SEOFixed — Per-page metadata + self-referencing canonicals

What this run is — and is not

Read this before you quote the number

  • It is a 3-agent free audit, not the full 20-agent suite. A full suite takes longer and checks far more.
  • The one “critical” in the count was our own scanner erroring (require is not defined), not a defect on the page. We logged it, fixed the scanner, and left it in the tally rather than quietly deleting it.
  • This is a point-in-time snapshot from 31 March 2026. Several of these findings have since been fixed. Re-run it yourself and compare.
  • A pass rate measures the site being scanned, not the scanner. 61.8% is a statement about p1qa.com, not about P1·QA's accuracy.

Why there are no customer case studies here

Because we do not have published customer results yet. Writing three plausible-sounding companies with plausible-sounding executives saying plausible-sounding things would be faster than this page. It would also be a lie, and you would be trusting us to find lies in your software.

So instead: our own scan, our own bad grade, the raw artifact name, and the caveats. When a customer agrees to be named with real numbers, those numbers will appear here with the same provenance.

20 specialized agentsFree audit measured at 66.5sSelf-scanned on every push, PR, and daily

Run the same scan on your app

Same three agents, same checks, same report format. No signup. You will get a grade, and it might be worse than ours.

Free QA Audit