EasyTaskAI — quality report QR-26-00088 | Vibe96

‹ EasyTaskAIQuality report QR-26-00088

EasyTaskAI
Vibe96 quality report

EasyTaskAI

AI-powered social media manager for creators, agencies and small businesses

Product ID
B0308-98564
Builder
Jehan Ahmed
Version reviewed
1.5.0 · released 4 Nov 2026
Stack
Express 5.3 · Node.js 20+ · PostgreSQL 15+ (Supabase)
Built with
Lovable
Review period
30 Oct 2026 – 4 Nov 2026
77out of 100 Passed with 4 warningsPass mark 75 and no failing check.

Scores

Security75
Code structure72
Docs & completeness73
Dependency health82
Build & deploy82
  1. AI code audit6 checks · 30 Oct 2026
  2. Installed & run3 Nov 2026
  3. Demo vs delivered4 Nov 2026
  4. Final sign-off4 Nov 2026

At a glance

  • Installed and run from the delivered code — 8 of 8 checks answered yes.
  • Live demo compared with the delivered code — 7 of 7 checks answered yes.
  • 11 of 15 promised items delivered (setup guide, example config and the rest).
  • Security scored 75/100 with no failing check.
  • 4 open warnings — explained in part 1.
Part 1

AI code audit

The full source archive was read by our AI auditor against each check. It reads the code — it does not run it; running it is part 2.

CheckWhat we foundScoreResult
Secrets & credentialsEvery secret-scan candidate is either a variable reference (process.env, cfg.supabaseAnonKey) or OAuth token handling at runtime, not a hardcoded credential. .env.example holds only blank placeholders, and the README states demo passwords are not shipped. Small deductions for items that could not be fully verified from the excerpts.92Pass
Dependency vulnerabilitiespackage.json pins all 10 production dependencies to exact versions and package-lock.json exists, so there are no wildcards, missing lockfiles, or known critical vulnerabilities that I can confirm. A few versions are unusually new, so I cannot verify them against advisories from the digest alone. The Node version requirement is inconsistent across package.json, docker-compose and the declared stack.82Pass
Authorisation & access controlOnly a fraction of server.js (223 KB) was visible, so the route-by-route trace could not be completed. What is visible is encouraging: a requireAuth middleware, a resolveWorkspace membership check, owner-scoped service-role queries, and signature verification on the Paddle webhook. However, the service-role client bypasses RLS, so a single missed ownership check exposes data, and several public endpoints need confirmation.60Warning
Database access rulesThe visible code uses supabase-js (parameterized) with a service-role client and relies on code-level owner scoping, with RLS as defense-in-depth. No injection path or unscoped delete is visible in the provided excerpts. Most of server.js (223 KB) and the 37 migrations were not shown, so scoping, foreign keys and indexes cannot be fully verified, and the service-role pattern adds risk.72Warning
Code structure & duplicationThe lib/platforms, lib/video and worker split is sensible and each platform adapter follows one interface. However, server.js (223 KB) holds all /api routes, and a project file mixes theme, nav, auth and chrome concerns, which makes the codebase hard to navigate for a buyer extending it.72Warning
Build reproducibilityA clean machine can most likely restore and start the app: package.json pins exact versions, engines requires Node >=22, the README orders the steps database, config, then processes, and docker-compose is coherent. The declared stack (Node 20+) conflicts with the real requirement (Node 22+), and several referenced files were not shown. Those gaps stop this short of a clean pass.82Warning

Open warnings

Problems that did not stop the listing. They stay on the product page until a later version fixes them.

Warning 1 Structure is modular per platform, but server.js is a 223 KB god file and i18n.js is 207 KB

The lib/platforms, lib/video and worker split is sensible and each platform adapter follows one interface. However, server.js (223 KB) holds all /api routes, and a project file mixes theme, nav, auth and chrome concerns, which makes the codebase hard to navigate for a buyer extending it.

Warning 2 DB access looks owner-scoped by design, but the excerpts don't show the queries or the schema, so key gaps stay unverified

The visible code uses supabase-js (parameterized) with a service-role client and relies on code-level owner scoping, with RLS as defense-in-depth. No injection path or unscoped delete is visible in the provided excerpts. Most of server.js (223 KB) and the 37 migrations were not shown, so scoping, foreign keys and indexes cannot be fully verified, and the service-role pattern adds risk.

Warning 3 Mostly reproducible; Node version and stack declarations conflict, and the setup guide could not be fully verified

A clean machine can most likely restore and start the app: package.json pins exact versions, engines requires Node >=22, the README orders the steps database, config, then processes, and docker-compose is coherent. The declared stack (Node 20+) conflicts with the real requirement (Node 22+), and several referenced files were not shown. Those gaps stop this short of a clean pass.

Warning 4 Authz cannot be fully verified: routes mostly hidden; webhooks and demo login need review

Only a fraction of server.js (223 KB) was visible, so the route-by-route trace could not be completed. What is visible is encouraging: a requireAuth middleware, a resolveWorkspace membership check, owner-scoped service-role queries, and signature verification on the Paddle webhook. However, the service-role client bypasses RLS, so a single missed ownership check exposes data, and several public endpoints need confirmation.

Requirement list

What the builder agreed to deliver before uploading — 11 of 15 met.

  • Complete source for every component Met
  • No real credentials in the archive Met
  • Builds from a clean machine Partial
  • Database script or migrations Met
  • Authentication and roles on every mutating endpoint Partial
  • Pinned dependencies without critical CVEs (nice)Partial
  • Documentation set (nice)Met
  • Working live demo Partial
  • README.md Met
  • HowToSetup.md Met
  • CLAUDE.md (or AGENTS.md) (nice)Met
  • CHANGELOG.md (nice)Met
  • .env.example Met
  • a project file Met
  • a project file (nice)Met
Part 2

Human review

Reviewers from the Vibe96 team in Dubai worked through each question below; a final reviewer signed it off.

Stage 1 — Installed and run from the delivered code

Completed 3 Nov 2026

A reviewer installed the product on a clean machine, following only the guide that ships with it.

QuestionAnswerReviewer's note
Packages restored without errorsYes
Solution builds cleanYes
Database created from the shipped script or migrationsYes
First sign-in works with the documented stepsYes
Setup guide is accurate and completeYes
Every module in the feature list opens and worksYes
No real secrets in the archiveYes
Branding and company details can be changed without codeYes

Stage 2 — Live demo compared with the delivered code

Completed 4 Nov 2026

A second reviewer checked that what the listing and demo show is what the buyer receives.

QuestionAnswerReviewer's note
Demo URL is reachableYes
Demo credentials workYes
Demo runs the same version as the archiveYes
All five differentiators are visible in the demoYes
Screenshots match the demoYes
Nothing in the demo is missing from the codeYes
Datasheet claims are accurateYes

Changes asked for during review

  • None — approved on the first round.
  • Sign-off: approved for sale on 4 Nov 2026 with 4 warnings shown publicly.

What this report covers — and what it doesn't

  • It covers version 1.5.0 exactly as the buyer downloads it: the source archive, the setup guide and the live demo on the dates above.
  • It is not a penetration test, a performance or load test, or a legal review of third-party package licences.
  • Scores are a point in time. Libraries age; a later version is audited again and gets its own report.
  • Your setup matters. Hosting, configuration and changes made after purchase are outside this review.
  • How the score is made: each AI check is scored 0–100 and weighted (security counts most). A listing needs 75 or more, no failing check, and the human review passed.
  • Always try the live demo before buying, and ask the builder questions from the product page.

Report history

ReportVersionDateScoreStatus
QR-26-000881.5.04 Nov 202677Current
QR-26-000671.0.027 Oct 202692Replaced

Vibe96 FZ-LLC, Dubai, UAE. This report describes the product at the time of review and is not a warranty.