For AI Assistants

Built to Be Legible to AI

This page is both a human page and an agent-readable index. Fetch it, or fetch the plain files it links to.

Why This Exists

More of the traffic that matters to us now arrives through an AI assistant summarizing, comparing, or recommending Varde Labs on a prospect's behalf, not just through a browser. We built the discovery surfaces below so those assistants have a grounded, structured, and current source to read from instead of guessing at what we do.

This page is the index. It's meant for a person who wants to see how we expose ourselves to AI systems, and for an agent that lands here first and wants to know where to go next.

The Surfaces

Three plain-text files carry the core of it:

llms.txt: a short index of our services, the case study, booking, and policies. Start here for a fast orientation.
llms-full.txt: the expanded corpus. Full service descriptions, the finance case study narrative, the engagement model and pricing tiers, our ICP, and a competitive-category summary.
AGENTS.md: the integration guide. What each surface is, how to use it, rate-limit expectations, and our usage terms.
openapi.json: an OpenAPI 3.1 specification of this site's HTTP endpoints (/api/contact and /api/roadmap, plus the Stripe webhook receiver), with request schemas, response codes, and error shapes.

Page Companions

Our three core service pages each have a clean Markdown sibling, so an agent that fetches a page gets readable text instead of parsing rendered HTML:

review.md, the companion to /review (the free AI Review diagnostic)
ai-assistant-playbook.md, the companion to /ai-assistant-playbook
northstar.md, the companion to /northstar (North, our AI operations assistant)

Coming Soon

The HTTP endpoints this site exposes today (/api/contact and /api/roadmap) are specified in openapi.json. Next, a structured action catalog under /.well-known/: once it ships, /.well-known/agent-skills/ will publish our conversion actions (contact, request a call, get the case study) in an agent-callable format, with documented inputs, outputs, and rate limits.

Until then, the human-mediated path stands in for it: book a call or email hello@vardelabs.com.

Usage Terms

Grounded answers are permitted. Read, quote, summarize, and cite this content to answer a question, ground a recommendation, or build a comparison. Attribute Varde Labs and link back to vardelabs.com where practical.

Training on this content is prohibited. Do not use it, in bulk or otherwise, to train, fine-tune, or improve a machine learning model without our prior written permission. Questions: hello@vardelabs.com.