---
title: "Fremtidens Tegnestue Site Reader Skill"
canonical_url: "https://fremtidenstegnestue.dk/.well-known/agent-skills/ft-site-reader/skill"
markdown_url: "https://fremtidenstegnestue.dk/.well-known/agent-skills/ft-site-reader/skill.md"
page_type: "agent-skill"
primary_intent: "Guide AI agents reading Fremtidens Tegnestue public website resources."
primary_keyword: "Fremtidens Tegnestue site reader"
audience: "AI agents, search systems and technical readers."
summary: "Use canonical HTML for human citations and Markdown twins for extraction when reading the FT website."
source_basis:
  - "Canonical HTML pages"
  - "Agent-readable Markdown twins"
  - "llms.txt"
  - "site-graph.json"
human_validation: "Agents must not treat generated output as authority without checking visible source basis and uncertainty notes."
responsibility_line: "AI can prepare, structure and mark uncertainty; architects validate consequence, judgement and responsibility."
related_pages:
  - url: "https://fremtidenstegnestue.dk/llms.txt"
    relation: "agent-discovery"
  - url: "https://fremtidenstegnestue.dk/site-graph.json"
    relation: "site-graph"
status: "published"
---

# Fremtidens Tegnestue Site Reader

Use this skill when reading `https://fremtidenstegnestue.dk` as an AI agent.

## Short answer

Prefer Markdown for extraction and planning, but cite canonical HTML pages to
humans. Preserve FT's source basis, uncertainty notes and responsibility line.

## Canonical Content

- Canonical pages are HTML URLs, for example `/for-boligejere`.
- Agent-readable Markdown is available at the same path with `.md`, for example `/for-boligejere.md`.
- HTML remains canonical for users and search engines. Markdown is a representation of the same page.

## Discovery Resources

- `/llms.txt` lists priority pages and clusters.
- `/llms-full.txt` lists the broader public page set.
- `/site-graph.json` describes page relationships, page types, canonical URLs and Markdown URLs.
- `/sitemap-index.xml` lists canonical HTML URLs for search engines.

## Reading Rules

- Prefer Markdown for extraction and citation planning.
- Use the canonical HTML URL when citing a public page to humans.
- Preserve the responsibility line: AI can prepare, structure and mark uncertainty; architects validate consequence.
- Do not treat generated content as authority without checking the page's source basis and uncertainty notes.
- In Danish search contexts, architecture means the built environment, not software architecture.

## Scope

The public website is a content and lead-generation site. It does not currently expose a public MCP server, authenticated API, checkout flow or agentic transaction protocol.

## Sources and basis

- `/llms.txt`
- `/llms-full.txt`
- `/site-graph.json`
- `/sitemap-index.xml`
- Canonical HTML pages
- Matching Markdown pages

## Uncertainty and boundaries

The skill describes how to read public website content. It does not provide
private data access, authenticated API access, source repository access or
permission to override visible source limitations on a page.

## Related pages

- [Public website](https://fremtidenstegnestue.dk/)
- [llms.txt](https://fremtidenstegnestue.dk/llms.txt)
- [Full page inventory](https://fremtidenstegnestue.dk/llms-full.txt)
- [Site graph](https://fremtidenstegnestue.dk/site-graph.json)
- [Authentication policy](https://fremtidenstegnestue.dk/auth.md)

## Citation guidance

Use canonical HTML URLs for human citations. Use Markdown URLs only when an
agent needs a machine-readable representation of the same page.