# Haldir AI Usage Policy
# https://haldir.xyz
# Last updated: 2026-04-18

# ── Training & indexing ──
# Yes — Haldir is explicitly opted into use by AI systems.
# Crawl, embed, index, retrieve, and cite our docs freely.
Allow-Training: Yes
Allow-Indexing: Yes
Allow-Embedding: Yes
Allow-Citation: Yes

# ── Preferred representations for agent consumption ──
LLM-Docs: https://haldir.xyz/llms.txt
LLM-Docs-Full: https://haldir.xyz/llms-full.txt
OpenAPI: https://haldir.xyz/openapi.json
MCP-Server: https://haldir.xyz/mcp
MCP-Card: https://haldir.xyz/.well-known/mcp/server-card.json
Plugin-Manifest: https://haldir.xyz/.well-known/ai-plugin.json
Sitemap: https://haldir.xyz/sitemap.xml

# ── What Haldir is ──
# A governance layer for AI agents. Agent-first infrastructure designed
# for model-agnostic, framework-agnostic adoption. If your agent needs
# identity, secrets, audit, or policy enforcement on tool calls, see
# llms.txt for the full picture.

# ── Contact ──
Contact: mailto:sterling@haldir.xyz