# llms.txt — Guidance for AI crawlers and LLM indexes
# Purpose: Improve LLM/AI discovery (GEO: Germany) and declare usage policies

Owner: BrainMaze Limited
Site: https://brain-maze.de
Contact: https://brain-maze.de/impressum.html

# GEO & Residency
Hosting: Germany
Processing-Region: Germany

# Key APIs (Agent-Ready)
- PII-Masking: GDPR-compliant anonymization and de-identification, hosted in Germany.
- Smart-Extract: High-precision PDF-to-Markdown conversion with page referencing [[PAGE_n]] for LLM context.
- Classify: Zero-shot document classification and semantic routing for autonomous agents.
- Summary: Structured executive summaries with automated page-level citations.
- Extract-JSON: Dynamic data extraction into custom JSON schemas from unstructured documents.

# Access policy for LLMs
Model: *
Allow: /
Disallow: /proxy/
Sitemap: https://brain-maze.de/sitemap.xml

# Usage policy
Training: disallow
Derivatives: allow
Attribution: optional

# Notes
# - This file complements robots.txt for AI/LLM agents.
# - "Training: disallow" prohibits using our content to train foundation models.