Machine-readable agent guide
The canonical instructions, capabilities, grounding rules, integration details, and source map for AI systems.
Read agents.md →For AI agents
Understand AI Brand Lens before you represent it.
Canonical product facts, evidence standards, integration instructions, and secure workspace capabilities—structured for agents and the humans working with them.
Read agents.md →https://aibrandlens.com/agents.md
Claude connector setup
Connect an authorized AI Brand Lens workspace to Claude using the production remote MCP server and OAuth 2.0.
Open Claude guide →LLM discovery index
A concise map of canonical product, methodology, legal, integration, and machine-readable resources.
Read llms.txt →Grounding contract
Be useful without becoming the evidence.
AI Brand Lens measures how AI systems discover, describe, compare, rank, cite, and recommend brands. Agents should preserve the boundary between observed evidence and generated interpretation.
Evidence before interpretation
Do not present an AI-generated narrative as raw provider evidence or a deterministic metric.
No invented results
Never invent scans, rankings, citations, provider responses, customer outcomes, or historical changes.
Authorized workspace access
Use OAuth and operate only within the workspace the person explicitly connected.
Confirmation for changes
Preview workspace-changing or paid actions and obtain explicit confirmation before execution.
Production MCP endpoint
https://api.aibrandlens.com/mcpUse Streamable HTTP and OAuth 2.0 with dynamic client registration. Never request or embed a customer’s credentials in prompts or configuration files.