/* Adjust small text above subpage headers to match freysa.ai */
#header .eyebrow {
    /* @apply text-gray-500; */
    color: var(--freysa-text-secondary);
    text-transform: uppercase;
    font-weight: 600;
    font-size: 0.875rem; // 14px
}

#topbar-cta-button > a > span {
    /* matches rounded-xl so top nav bar CTA button has the same roundedness as the search input */
    border-radius: .75rem; 
}

/* ! Override link colors in page content *only* when not in dark mode. Otherwise they default to the primary-light color, which is basically white and invisible in light mode. */ 
html:not(.dark) .prose :where(a):not(:where([class~=not-prose],[class~=not-prose] *)) {
    /* color: var(--tw-prose-links); */
    /* text-decoration: none; */
    /* font-weight: 600; */
    /* --text-primary-light: #f00 !important; */
    border-color: rgb(var(--primary-dark)) !important;
}

// This prevents a new window from opening when clicking a freysa link.
document.querySelectorAll('a[href*="freysa.ai"]').forEach(link => {
    link.setAttribute('target', '_self');
    link.removeAttribute('rel');
});

Building with this stack

Freysa Framework

Freysa is the first sovereign AI agent you can integrate with.

Overview

Roadmap: The Seed

Introduction

Attestation endpoints return cryptographically signed evidence about the code and configuration running in the Trusted Execution Environment (TEE). This evidence allows third parties to verify the authenticity and integrity of the TEE's execution environment.

Get attestation details

OpenRouter/OpenAI API-compatible completions endpoint enables developers to build their own applications on sovereign agents. The completions request object is augmented with appropriate details from the agent's memory. Following the standard API format ensures compatibility with all SDKs that support OpenRouter or OpenAI API.

Create chat completion

Memory retrieval endpoint enables querying of the agent's memory to obtain the top-k most relevant results.

Query memories

Memory endpoint enables ingestion of memories into your agent. These memories are used by the agent when using tools or when developers invoke the chat completions endpoint.