# developer-ai.txt for BSolve IT Limited
# Spec: https://www.ai-visibility.org.uk/specifications/developer-ai-txt/ v1.8.0
# Last-Updated: 2026-07-26Lang: en-GB

[overview]
BSolve IT Limited is the registered parent company of five UK web brands, based in
Kettering, Northamptonshire. The website (https://www.bsolveit.com/) is a hand-coded,
server-rendered PHP site of 16 pages: an entity hub identifying the group, a directory of the five
brands, and the one place all of the group's work is shown together.
It has NO user accounts, NO e-commerce, NO public API, NO analytics, NO tracking and NO advertising.
It sets no cookies of its own. There is ONE form, on /contact/, and ONE third-party script behind it:
Google reCAPTCHA v3, fetched on first interaction with that form rather than on page load. Everything
else, fonts included, is served from this origin.
It replaced a 475-URL WordPress site in July 2026; 449 of those URLs now return HTTP 410 Gone.

[public-api]
status: not-available
notes: BSolve IT Limited does not offer a public API. The contact form at /contact/
posts to /api/contact/, which is NOT a public API: it is reCAPTCHA-gated, rate limited, accepts a
fixed set of four subjects and returns only {ok:true} or an error code. Email reaches the same
inbox. The closest thing to a machine-readable interface is this AI Discovery
suite plus /sitemap.xml, and every published figure carries its source URL and measurement date
in the page's JSON-LD (Dataset nodes) as well as in the visible text.

[public-areas]
- / - entity hub. The hero is a showreel: one live group site at a time inside browser chrome carrying its real domain, advancing every 6s with an explicit labelled pause control, and a list of the eight domains beside it
- /about/ - the company, its history since December 2002, and what it does not do
- /brands/ and /brands/<slug>/ - the five brands, one page each
- /work/ - every live site the group builds, hosts and runs, with measured outcomes
- /standard/ - the build standard every site in the group has to meet
- /people/mark-mcneece/ - author entity page
- /contact/, /accessibility/, /editorial-standards/, /privacy-policy/, /terms/
- /410.php - the branded Gone page served by the 449 retired URLs
- AI Discovery Files - /llms.txt, /llms.html, /ai.txt, /ai.json, /identity.json, /brand.txt, /faq-ai.txt, /robots-ai.txt, /developer-ai.txt (this file). /llm.txt 301s to /llms.txt, which is the behaviour the specification defines
- /.well-known/security.txt - RFC 9116 disclosure contact
- Sitemap at /sitemap.xml, indexing /page-sitemap.xml
- /assets/ - static images, CSS, fonts, JavaScript

[authentication]
None. The site has no user accounts and no authenticated areas.

[technology-stack]
- Server: PHP 8.x on Apache (365i shared hosting, StackCDN edge cache via 20i)
- Frontend: HTML5, vanilla CSS, vanilla JavaScript (no framework, no bundler)
- Build step: optional CSS/JS minification. Live cache-busting via bsit_asset() appending ?v=<filemtime>
- All page content rendered from PHP data arrays in includes/data/ so visible content and JSON-LD schema cannot drift
- One form (/contact/ posting to /api/contact/): reCAPTCHA v3 gated, honeypot, header-injection guards, an allowlist of four subjects, rate limited to 5 per hour per IP counted only after a successful send. It sends TWO messages over SMTP, a notification to the office and an acknowledgement to the sender, both built by includes/email.php as HTML with a plain-text alternative. Credentials live in an untracked includes/secrets.env, never in the repository
- The failure mode throughout is CLOSED: no reCAPTCHA secret means no submissions accepted, never silent acceptance
- Motion layer: self-hosted GSAP + ScrollTrigger, loaded only on pages that opt in via $page_motion
- Typography: three self-hosted woff2 files totalling 95 KB (Switzer, Instrument Sans, Martian Mono), all variable, all subsetted to latin
- URL dispositions are generated from data/disposition.csv into .htaccess rules and verified by a harness that asserts all 475 old URLs return their intended status
- Deploy: FTPS mirror of website/ followed by a StackCDN cache purge

[documentation]
- [Site (canonical)](https://www.bsolveit.com/)
- [llms.txt - LLM-readable site summary](https://www.bsolveit.com/llms.txt)
- [llms.html - human-readable rendering of llms.txt](https://www.bsolveit.com/llms.html)
- [identity.json - canonical identity data](https://www.bsolveit.com/identity.json)
- [faq-ai.txt - consolidated FAQ corpus](https://www.bsolveit.com/faq-ai.txt)
- [sitemap.xml - canonical URL index](https://www.bsolveit.com/sitemap.xml)
- [Contact page](https://www.bsolveit.com/contact/)

[technical-contact]
- [Contact form](https://www.bsolveit.com/contact/)
- Email: support@bsolveit.com
- Phone: +44 7785 326603


[metadata]
version: 1.8.0
last_updated: 2026-07-26