Changelog

What we shipped,
in the order it shipped.

AutoInk is in private beta. Every entry below is a real change you'd notice if you were already on the waitlist. Internal cleanups, refactors, and test coverage live in git log, not here.

Marketing

Pricing detail page and legal pages

The marketing site grows up: pricing has its own URL with a full feature matrix, and Privacy / Terms / Security are now public-readable.

  • /pricing — credit economy explainer, per-action cost table, side-by-side plan comparison, and dedicated FAQ
  • /privacy and /terms — v1 prose covering subprocessors, LLM data flow, retention, AUP, and refund policy
  • /security — encryption stack, audit findings, and responsible-disclosure address
  • /changelog — this page
BillingOperations

Billing gate enforcement is live

Free-tier workspaces are now metered against their credit allowance. Daily reconciliation watches the ledger against provider invoices so prices stay honest.

  • Reserve-and-settle gate: runs that would go negative are blocked before the LLM call fires
  • Per-workspace spend cap protects against runaway autopublish loops
  • Daily COGS reconciliation across Anthropic, OpenAI, Google, DeepSeek, Perplexity, xAI, DataForSEO, and Stripe
  • Drift over 5 % triggers an operator alert; we pause pricing changes until it's resolved
WorkspacesBilling

Team workspaces, plan tiers, and Stripe

AutoInk goes from solo tool to team product. Five plans, monthly or annual, with shared billing and seat-based access.

  • Team workspaces — invite by email, owner / member roles, workspace switcher, leave-and-return semantics
  • Five plans live (Free / Solo / Pro / Business / Enterprise) with credit allowances and per-tier feature gating
  • Stripe checkout for subscriptions; customer portal for upgrades, downgrades, payment-method changes, and cancellation
  • Annual billing offers ~17 % off the monthly rate
  • Top-up packs ($10 / $25 / $50) for paid plans
Console

Console redesign, second pass

A typography-first overhaul of the in-app console: tighter density, a brand-anchored serif lock, and a credit meter you can see at a glance.

  • New type ramp (display / heading / body / label / kicker / mono) applied site-wide
  • Iowan serif locked to brand anchors only — sidebar wordmark, auth, onboarding hero
  • Sidebar footer now shows live credit balance and cycle window
  • Main bundle dropped from 1.5 MB to 327 KB; OG editor and chart libs lazy-load on demand
Security

Internal security review closed

Before billing went live we ran a full backend + frontend review against the OWASP top 10 and our own checklist. Everything found was remediated in the same release.

  • SSRF defence on every outbound HTTP request, including redirect chains
  • Command injection eliminated by switching all hooks from string commands to argv arrays
  • Three-bucket login throttling (per-pair, per-IP, per-email) survives credential-stuffing botnets
  • Sessions are atomically rotated on password change
  • Argon2id passwords; AES-256-GCM credential encryption; HIBP k-anonymity check on signup
  • Live security posture at /security
GEO

GEO judge — second-pass citation detection

Structured citation scans miss 20–40 % of real engine usage according to industry research. AutoInk now runs a cheap LLM judge on every "not cited" result to flag content the engine used without naming.

  • New citation badge: Cited / Mentioned (est.) / Not cited / Not scanned
  • Judge uses Gemini Flash by default — adds about $0.0002 per article-engine pair
  • Inspector shows the judge's rationale alongside the structured scan output
GEO

Generative engine optimization tracking

AutoInk now asks ChatGPT, Claude, Gemini, Perplexity, Grok, and Google AI Overviews whether they're citing your articles, and records the answer every day.

  • Per-article timeline of citations across all six engines
  • Project-level breakdown — which engine surfaces you most, which one ignores you
  • Manual scan trigger and per-project auto-scan scheduler with cost cap
  • First scan fires three days after publish; weekly thereafter
Pipeline

Real-time SEO score and review-then-publish

Drafts now ship with an inline 0–100 SEO score and three publish modes per project.

  • SEO score breaks down into structure, keyword coverage, internal links, and length
  • Per-project publish modes: auto (ship the draft), review (queue for human approval), manual (save and stop)
  • Pending-review queue with batch publish
Pipeline

SERP reverse analysis

Before drafting, AutoInk now structurally analyses the top-ranking pages for the keyword and shapes the outline against what's currently winning.

  • DataForSEO live SERP fetch for the target keyword
  • Heading and section pattern extraction across the top 10
  • Outline auto-mirrored, then editable before drafting kicks off
Publishing

Multi-CMS publishing

Drafts can now be published directly from the run into seven destinations.

  • GitHub (commits to your repo via GitHub App)
  • WordPress, Ghost, Notion, Webflow, Framer, Payload
  • Credentials stored AES-256-GCM encrypted, decrypted only at the moment of publish
  • Test-connection probe before saving any new adapter config
Launch

AutoInk enters private beta

Day one. The full pipeline — keyword discovery, AI drafting, quality gates, publish — runs end-to-end.

  • Keyword discovery with intent classification
  • AI drafting on DeepSeek, Claude, Gemini, and OpenAI models
  • Quality gates (minimum word count, minimum SEO score) hold weak drafts before publish
  • Cost tracking per run with daily caps

Subscribe via Writing or email — we'll send a monthly digest of the bigger releases.

Join the AutoInk waitlist

We’re in private beta and onboarding in waves. We’ll email you as soon as a slot opens — usually within a few weeks.

We’ll only use this email to let you know when your slot opens. No newsletter, no sharing.