Trust / Security

Observe freely. Control deliberately.

Understand the public site’s read-only boundary and a defensive approach to authorization, secrets, data contracts, and production collector design.

What this public site does

The site serves static pages, original images, and fictional public data files. Its interactive controls filter samples, switch authored chart windows, reveal FAQ answers, copy visible text, and read bundled JSON. They do not connect to equipment or issue mining commands.

There are no account sessions, wallet connections, payments, log uploads, or API-key provisioning. The API reference lists the actual sample resources rather than implying a separate production service.

Keep private access out of public clients

Do not put device credentials or privileged tokens into HTML, JavaScript, published configuration files, screenshots, or downloadable examples. A public static client cannot keep those values secret from someone who can download it.

A real collector needs a separately secured environment and appropriate access to its upstream systems. The secure mining API guide provides a defensive design checklist grounded in established REST guidance.

Give read-only work read-only permissions

Monitoring should not automatically inherit authority to change pools, firmware, operating profiles, or other device settings. Where the upstream supports it, separate observation identities from privileged operations. Document limitations when a device cannot provide that distinction.

Production controls require explicit authorization, an appropriate review process, audit records, and a recovery plan. A generic “optimize” button is not a substitute for that operational design.

Treat AI inputs and outputs carefully

The public prompt template is not connected to a model. In a separate implementation, sanitize inputs, treat log content as untrusted, constrain privileges outside the model, validate outputs, and require human verification for important claims.

See the AI log-analysis guide for a bounded evaluation workflow. Prompt instructions alone are not a security boundary.

Report a concern without sending secrets

Use [email protected] to describe a suspected content or site issue. Include the affected page and a minimal explanation. Do not send credentials, private keys, raw sensitive logs, or information you are not authorized to share.

No response time, managed security service, or certification is promised. This page describes the shipped website's behavior and design boundaries rather than claiming a comprehensive security guarantee.

Your next idea starts with a little silicon.

Explore the guides. Inspect the data. Build with a clearer picture.