Everything Forge gives you

Complete list of what Forge generates and includes automatically. Every item below is active the moment you call app.Content(forge.NewModule(...)).

Routes and feeds

List route — GET /{prefix} returns all Published items as JSON or HTML
Detail route — GET /{prefix}/{slug} returns a single Published item
RSS feed — GET /{prefix}/feed.xml plus aggregate GET /feed.xml
Sitemap entries — GET /{prefix}/sitemap.xml merged into GET /sitemap.xml; regenerated on every publish, update, and delete
SEO meta tags — <title>, <meta description>, canonical URL in every <head>
Open Graph tags — og:title, og:image, og:type, article:published_time and more
Twitter Cards — twitter:card, twitter:title, twitter:image
AI index formats — /llms.txt compact index, /llms-full.txt Markdown corpus, per-item /aidoc token-efficient endpoint
Content negotiation — one URL returns HTML to browsers, JSON to APIs, and AI-optimised format to agents; no extra code

Database table — derived from your struct; works with PostgreSQL, SQLite, and MySQL
SQLRepo — production SQL repository with automatic table naming and upserts
SeqRepository — lazy streaming interface for large datasets (iter.Seq2)

Markdown rendering (forge_markdown) — including HTML passthrough for trusted blocks
Trusted HTML (forge_html) — verbatim emission of pre-rendered HTML fields
Field semantics in MCP schema — forge_format and forge_description struct tags; AI agents understand field intent without extra prompting

Draft / Scheduled / Published / Archived enforcement — hardwired, cannot be disabled
Scheduled publishing — automatic Scheduled → Published transition at ScheduledAt; no external cron
404 on everything non-Published — guests, search engines, and AI crawlers see nothing until explicitly published

Role-based access — Guest → Author → Editor → Admin enforced per module per operation
Struct-tag validation — forge:"required,min=3" enforced identically for HTTP, API, and MCP calls
Token management — named revocable tokens with role scoping; ensureBootstrap auto-creates the first admin token on first start
ErrLastAdmin guard — cannot revoke the last admin token

NavTree — first-class navigation abstraction (NavModeDB / NavModeCode)
4 MCP nav tools — list_nav_items, create_nav_item, update_nav_item, delete_nav_item

Per content type, automatically derived — no manual definition:

Shared partials — App.Partials + MustParseTemplate
HeadAssets — favicons, stylesheets, preconnect, scripts injected via forge:head on every page
ContextFunc — per-request extra data passed to module templates
Static file serving — App.Static serves from embedded FS in production (immutable cache headers) and from disk in development

JSON-LD — Article, Product, FAQ, HowTo, Event, Recipe, Review, Organisation rich results
OGDefaults — site-wide fallback OG image and Twitter handles
AppSchema — site-wide Organisation/WebSite JSON-LD on every page
Robots — <meta name="robots"> set per lifecycle status; configurable AI crawler policy

File-based configuration — key = value format, fail-fast at startup
/_health endpoint — returns framework version and status; exempt from HTTPS redirect
Zero third-party dependencies — pure stdlib; driver is always your choice
Cookie compliance — /.well-known/cookies.json declares all cookies with category and consent requirements
Redirect tracking — automatic 301 on slug rename, 410 Gone on archive/delete

Typed MCP tools derived directly from your content type — no manual schema definition
Field semantics — AI agents understand what each field means without extra prompting
Agents operate under the same access rules as humans — no special bypass
forge-cli — init, content CRUD, token management, and media operations from the terminal
Go codebase designed to be readable and extensible by AI agents