Agent-Ready SaaS: The Machine-Readable Docs Answer Engines Actually Read

Generative engine optimization is not a keyword trick. It is whether an AI system can find a stable, quotable, machine-readable statement about your product without guessing.

The read order

1. llms.txt — a short, curated overview with links, written for an LLM with a limited context window.
2. llms-full.txt — the long-form facts: full configuration enums, pruning rules, API surface and honest limitations.
3. OpenAPI 3.1 — exact request and response schemas for every public route.
4. MCP endpoint — a stateless JSON-RPC tool an agent can call to obtain the current configuration and even a build.
5. Per-page JSON-LD — Organization, WebSite, SoftwareApplication/OfferCatalog, FAQPage and BlogPosting.

Robots should explicitly allow the machine-readable routes for AI crawlers, and the sitemap should list them. An FAQPage graph matters because answer engines quote question-and-answer pairs, not marketing paragraphs.

What belongs in the downloaded ZIP

The customer's generated project gets its own AGENTS.md (repo contract and definition of done), a style-specific design.md, scoped rules with globs, Cursor rules, CLAUDE.md and up to twelve skills playbooks — all rendered from one shared spec so they never drift from shipkit.config.json. Our site's llms.txt and MCP endpoint are site-only and must NOT leak into the customer package.

Structured content, not keyword stuffing

Every article here carries a visible FAQ section built from the same structured Q&A data that produces its FAQPage JSON-LD. The byline and publisher are the company name, not an invented person. Dates, canonical URLs, inLanguage and keywords are explicit. A Markdown variant of each article is published so a crawler can read the text without rendering the page.

Honest limitations are GEO assets

Vague claims get ignored; specific limits get quoted. State plainly that SMTP is not implemented, that the local auth path has no MFA or OAuth, that MAIL_PROVIDER defaults to none, that the external error sink is optional, that the Cloudflare target needs Supabase because node:sqlite has no writable disk on Workers, and that the refund window is a plain 14 days (delivery events are evidence, not a condition). A machine-readable limitation is more trustworthy than a paragraph of adjectives — and it must be updated the moment the code changes.

How to measure it

Keyword volume is the wrong metric for GEO. Track AI referral traffic in analytics, brand mentions and share of voice in AI answers, and whether the cited page is one of the machine-readable surfaces. If an answer engine cannot cite you, more landing pages will not fix it.