Website just launched on a $0 Budget, so ai model requests (includes tools and chats) are limited to 150 a day (for everyone using the site) due to openrouter free limits

Built on a security-first foundation

The Mind Security Engine adds security headers, local suspicious-request detection, upload validation, human verification for suspicious traffic, API throttling, and optional Arcjet protection. Supabase row-level security keeps signed-in users scoped to their own studios, files, chat history, and settings.

Security layers

What the Mind Security Engine protects

ScholarMind applies multiple overlapping layers of protection so you can focus on studying without worrying about data safety, abuse, or unexpected costs.

End-to-end encryption headers

Strict security headers (CSP, HSTS, X-Frame-Options) are applied to every response to prevent common web attacks.

Suspicious-request detection

Local pattern-matching flags unusual request behaviour before it reaches your study data.

Upload validation

Each uploaded file is checked for type, size, and content safety before it is stored or processed.

Human verification

Suspicious traffic triggers a lightweight human check — keeping bots out without a noisy CAPTCHA.

API throttling

Rate limits and hourly/daily AI usage caps prevent runaway costs and abuse of the study tools.

Arcjet optional protection

Arcjet can be enabled for deeper threat intelligence — bot detection, VPN blocking, and IP reputation.

Supabase RLS at rest

Row-level security in Supabase keeps signed-in users scoped to their own studios, files, chat history, and settings.

How it works

Request lifecycle through the engine

01

Request arrives

Every API request hits the Mind Security middleware before reaching your study data.

02

Headers are set

Security headers are injected into the response, locking down what browsers and scripts are allowed to do.

03

Behaviour is scored

The engine scores the request against known abuse patterns — unusual rate, unexpected payloads, bot signatures.

04

Decision is made

Clean requests pass through. Suspicious requests get a human challenge. Malicious requests are blocked entirely.

Transparency

What we monitor and what stays private

What is monitored

  • Request patterns — rate, origin, and behaviour scoring
  • Feature usage — aggregate trends for reliability
  • Upload validation — file safety before processing

What stays private

  • Your study sources, AI outputs, and chat history
  • Studio content is never exposed publicly
  • Personal data is accessible only via Supabase RLS
For full details on data handling, AI processing, and your controls, read the Privacy Policy. The host panel can view aggregate usage and onboarding trends. It is not designed to expose private study answers publicly.

Study with the security you deserve

ScholarMind is built from the ground up to keep your study data safe. The Mind Security Engine runs on every request.