For developers, agents & LLMs

Everything here is also data.

This site publishes its own facts through a small, static JSON API. It is public, read-only, and needs no authentication, no API key and no payment. Fetch anything below straight from your code, an agent runtime or curl.

Jump to the endpoints OpenAPI 3.1 spec→

Ground rules

// policy

Read-only, key-free, static.

No authentication. No keys, no accounts, no sign-up. Every URL on this page answers 200 to anyone.
Read-only. Only GET exists — there are no write operations anywhere in this API.
Static files. Responses are generated at build time and change only at deploy, so cache them freely; there is no rate limit beyond common sense on a shared host.
Additive stability. Payload shapes carry a schemaVersion; changes within a version are additive only.

The JSON API

// 6 endpoints · GET · no auth

Six endpoints, one discovery document.

Start at /api/index.json — it lists everything below and never goes stale. A formal,OpenAPI 3.1 specification of all six operations (unique operationIds, typed response schemas — suitable for LLM function calling) is served at /openapi.json.

/api/index.json

This discovery document: lists every JSON endpoint with a one-line description, plus related machine-readable surfaces (llms.txt, feeds, agent manifest, sitemap).

curl -s https://automancer.uk/api/index.json

/api/business.json

Canonical business facts: legal name, trading name, company registration, ICO registration, contact details, address and areas served.

curl -s https://automancer.uk/api/business.json

/api/services.json

Service offerings with published from-prices (GBP), billing units and stable ids — the same data the /services pricing page is generated from.

curl -s https://automancer.uk/api/services.json

/api/case-studies.json

Every published case study: title, summary, sector, tags, publication date, headline stats and the full body text as raw Markdown.

curl -s https://automancer.uk/api/case-studies.json

/api/field-notes.json

Every published field note article: title, description, author, date, category, tags and the full body text as raw Markdown.

curl -s https://automancer.uk/api/field-notes.json

/api/pages.json

Every page emitted by this site: canonical URL, title, description, type, last-modified date and its Markdown twin URL where one exists.

curl -s https://automancer.uk/api/pages.json

Markdown twins

// path + ".md"

Every content page has a Markdown twin.

Append .md to any content page's path for clean Markdown with YAML front matter (Content-Type: text/markdown; charset=utf-8). For example: https://automancer.uk/about.md, https://automancer.uk/services.md, https://automancer.uk/field-notes.md, and every article: https://automancer.uk/work/<slug>.md, https://automancer.uk/field-notes/<slug>.md. The bodies are derived from the same source as the HTML pages, not retyped. Legal pages (/privacy, /terms) deliberately have no twin — fetch their HTML for the authoritative text.

Full-text guides & feeds

// for humans, agents and crawlers

The whole site, in one file if you want it.

/llms.txt — short plain-text guide to the business and this site, written for LLMs and agents.
/llms-full.txt — the complete site text in one file, each section headed with its canonical URL.
/work/feed.json and /work/rss.xml — case studies, full content, not excerpts.
/field-notes/feed.json and /field-notes/rss.xml — field notes, likewise.
/sitemap-index.xml — every emitted page.

Agent manifest

// capabilities & permissions

One manifest describing how to treat this site.

/agent.json is a capability manifest: contact routes (and their caveats), every endpoint, crawling policy, permitted uses and the legal-pages stance. It is also emitted at /.well-known/agent.json for hosts that serve dot-paths. Security-contact details live in /security.txt (RFC 9116). Questions a manifest can't answer? Email waseem@automancer.uk — a human reads it.