Source connection and extraction
Point Atlas at a website, employee handbook, brand guide, or Notion workspace and it extracts company context from those sources.
Atlas
RivendicaAtlas turns company docs and website content into shared AI rules for Claude, Cursor, and ChatGPT to use the same company context.
Overview
Atlas is a company-context layer for AI tools. It takes the rules, process knowledge, data definitions, and brand guidance that live in a company’s documents and makes them available to tools like Claude, Cursor, ChatGPT, Claude Code, and Codex.
The product is designed for teams that want AI outputs to reflect how the company actually works instead of starting from a generic prompt every time. Atlas lets teams connect source materials, extract structured context, and serve those rules through MCP or REST so they can be used across different agents and internal tools.
Features
Three-layer company context
Atlas organizes context into process, data, and brand layers so the output covers operational instructions, schema and metric context, and visual identity.
Agent delivery through MCP and REST
The product serves rules to Claude, Cursor, ChatGPT, Claude Code, Codex, and other agents through a single MCP URL or REST endpoint.
Markdown-based rules export
Rules remain in plain markdown and can be exported at any time, which the source presents as a no-lock-in setup.
Separate personal and company context
A shared company layer is managed by admins, while personal rules stay on each user’s account and are not shared with teammates.
Dashboard and setup support
The pricing page notes a dashboard where users can view and edit rules, plus setup help from Atlas’s team on the paid plan.
Use cases
Standardizing internal workflows
A team can centralize how common requests should be handled so employees do not have to re-explain internal processes every time they ask an AI assistant for help.
Answering questions against company data context
A company can keep schema names and metric definitions available to AI tools so analysts and builders get answers that match the company’s data model.
Keeping AI-generated brand assets on-brand
Marketing, design, or go-to-market teams can load brand guidance so slide decks and visual outputs follow company colors, fonts, and identity conventions.
Feeding context into custom agents
Developers can connect Atlas through MCP or REST and use the extracted rules inside custom agents or internal applications.
Separating personal and shared instructions
Individual users can keep private rules on their own account while the company maintains a shared layer for team-wide guidance.
Pros and Cons
Pros
- Captures company-specific process, data, and brand context in one place.
- Supports both direct AI tools and custom builds through MCP and REST.
- Keeps personal context separate from shared company context.
- Exports rules to markdown, reducing reliance on a single system.
Cons
- The source does not spell out detailed setup requirements beyond connecting a website or documents and, optionally, Notion.
- Integration details are high level; the source names supported tools and interfaces but does not document specific per-app behavior.
- The paid plan mentions hands-on setup help, which suggests some teams may need assistance to get started.
FAQ
What does Atlas do?
Atlas connects a company’s own materials—such as a website, employee handbook, brand guide, and optional Notion content—and turns them into rules that AI tools can follow. The source also says the rules can be served through MCP or REST.
Is there a free plan?
The pricing page shows a free option for trying website rule extraction and viewing/editing rules in the dashboard, plus a Founding plan at $99 per company per month that adds document and PDF extraction, rule serving to AI tools, and hands-on setup help.
Who is Atlas for?
Atlas is positioned for teams where employees use tools like Claude, Cursor, and ChatGPT regularly and want the same company rules applied across those tools.
How does Atlas handle personal and company context?
The source says personal context stays on the user’s account, while company context is shared and curated by admins for the whole team.
Can the rules be exported later?
The source says Atlas can export rules to markdown and that there is no lock-in.
Quick Facts
- Category
- AI context layer for tools and agents
- Website
- atlas.nanonets.ai
- Core interface
- Dashboard plus MCP and REST delivery
- Supported sources
- Website, documents, employee handbook, brand guide, optional Notion
- Supported tools mentioned
- Claude, Cursor, ChatGPT, Claude Code, Codex