a# llm.txt - HD2.ai LLM Interaction Policy
# Version: 1.1
# Last-Updated: 15 January 2026
# Site: https://www.hd2.ai
# Contact: support@hd2.ai

########################################################################
# 0. PURPOSE
#
# This document describes the preferred behavior for Large Language
# Models, AI agents, and automated systems interacting with HD2.ai
# public content. It is intended as an intent and preference signal
# from HD2.ai — similar to robots.txt but specific to LLMs and AI.
########################################################################

########################################################################
# 1. IDENTITY & SCOPE
########################################################################

[1.1] SITE OWNER
- HD2.ai is a UK-based AI and development studio specializing in
  intelligent automation, private LLM systems, autonomous agents,
  and scalable AI infrastructure. :contentReference[oaicite:4]{index=4}

[1.2] SCOPE OF THIS POLICY
- Applies to:
  - https://www.hd2.ai/ and all public subpaths.
  - Specifically includes:
    - / (Home)
    - /what-we-do/ (Services)
    - /blogs/ (Blog/Articles)
    - /contacts/ (Contact forms & public info)
- Does NOT apply to:
  - Password-protected content
  - Any authenticated, internal, or client-only pages
  - APIs/endpoints requiring credentials

########################################################################
# 2. GENERAL BEHAVIOR EXPECTATIONS
########################################################################

[2.1] TRANSPARENCY
- Agents should:
  - Clearly disclose when responses rely on hd2.ai content.
  - Cite or link back to relevant pages when feasible.

[2.2] POLITENESS
- Respect existing crawl limits (robots.txt, HTTP limits).
- Do not overwhelm hd2.ai servers with high-frequency requests.

[2.3] REPRESENTATION
- Do NOT:
  - Imply official endorsement by HD2.ai unless explicitly authorized.
  - Claim to be an HD2.ai proprietary agent or internal tool.

########################################################################
# 3. CONTENT ACCESS & INDEXING
########################################################################

[3.1] PUBLICLY AVAILABLE CONTENT
- The following kinds of public content MAY be indexed and used
  for generating answers, with proper attribution:
  - **Home page** — company overview & mission :contentReference[oaicite:5]{index=5}
  - **What We Do** — services & solutions descriptions :contentReference[oaicite:6]{index=6}
  - **Blogs** — articles and thought leadership content :contentReference[oaicite:7]{index=7}
  - **Contact/Connect** — publicly displayed contacts & forms :contentReference[oaicite:8]{index=8}

[3.2] PREFERRED USES
- You MAY:
  - Summarize public pages with context and attribution.
  - Answer user questions about HD2.ai offerings using public text.

[3.3] RESTRICTED OR SENSITIVE AREAS
- You MUST NOT use content that is:
  - Behind login walls or protected by credentials.
  - From internal dashboards, admin tooling, private APIs,
    or client workspaces.

########################################################################
# 4. TRAINING & MODEL USAGE
########################################################################

[4.1] GENERAL TRAINING GUIDELINES
- HD2.ai prefers that:
  - Public hd2.ai content is **NOT used to train general-purpose
    LLMs** without explicit permission.
  - You MAY process content temporarily to generate answers.

[4.2] PRIVATE / CONFIDENTIAL DATA
- Data originating from:
  - Private clients
  - Authenticated HD2.ai instances
  Must be treated as confidential and NOT be used for general training.

[4.3] DERIVATIVE KNOWLEDGE
- You MAY:
  - Learn general concepts from public explanations.
- You MUST:
  - Avoid replicating proprietary frameworks verbatim beyond fair use.

########################################################################
# 5. SAFETY & MISUSE
########################################################################

[5.1] PROHIBITED USES
- Do NOT use hd2.ai content to:
  - Facilitate unlawful activity
  - Build competing products that recreate HD2.ai’s internal
    proprietary models, prompts, or systems without consent.

[5.2] ALIGNMENT SIGNAL
- Use HD2.ai content responsibly:
  - Respect privacy, legal standards, and client confidentiality.

########################################################################
# 6. OVERRIDES & FIRST-PARTY AGENTS
########################################################################

[6.1] FIRST-PARTY AGENTS
- HD2.ai may operate its own agents or processes that override
  this file under specific internal policies or contracts.

[6.2] CONTRACTUAL AGREEMENTS
- Any explicit, written agreement with HD2.ai takes precedence
  over this default llm.txt.

[6.3] VERSIONING
- LLM clients should periodically re-fetch this file to stay up to date.

########################################################################
# 7. CONTACT
########################################################################

[7.1] QUESTIONS / PERMISSIONS
- For data-use exceptions or partnerships:
  Email: support@hd2.ai
  Subject: “LLM Integration with HD2.ai”

# End of llm.txt