About AI Stack DB

Pre-publication draft. Replace bracketed placeholders before publishing. See content/site-pages/README.md for the placeholder list.

What this site is

AI Stack DB, published at https://aistackdb.com/, is an independent reference site about AI and SaaS productivity tools. It exists to answer one practical question: given the tool a reader is considering, what does it actually do, who is it for, what does it cost, and what are the credible alternatives?

The site is built around two content surfaces:

Tool pages — one structured page per tool, covering what the tool does, primary use cases, intended audience, pricing model and free-plan status, supported platforms, alternatives, caveats, and a clearly listed set of official sources.
Comparison pages — head-to-head pages that contrast two tools across the same dimensions and end with decision rules rather than a winner declaration.

The current editorial scope is the AI assistant, AI writing, AI research, AI productivity, AI coding, AI design, AI video, and AI voice categories. The category list is maintained in data/categories.json in this project's repository and is not changed without an editorial reason.

What this site is not

It is not a vendor, reseller, or system integrator. AI Stack DB does not sell the tools it writes about and does not implement them on a reader's behalf.
It is not a review aggregator. Pages are author-written from primary sources, not synthesized from third-party review scores.
It is not a news site. It does not chase release-day coverage; pages are updated on a freshness cadence described in editorial-policy.md.
It is not a sponsored marketplace. The default state for every page is no affiliate links and no sponsored placement; any future deviation is disclosed on the page where it applies.

Who runs it

AI Stack DB is operated by 루모메이트(LumoMate), based in South Korea. The operator is responsible for editorial decisions, source verification, and corrections. Where AI tools are used during research or drafting, that use is disclosed in the ai-disclosure-source-policy.md page and, where material, on the individual page.

For contact, corrections, takedown notices, or rights-holder requests, see the Contact page.

Editorial principles in short

The full standard is in Editorial Policy. At a glance:

Every factual claim about a tool — what it does, what plans it offers, what platforms it runs on — must trace to an official vendor source recorded in data/sources.json. Sources whose access status is needs_verification or blocked are not quoted as fact; the page directs the reader to verify on the vendor's official site.
Pages avoid quoting specific prices or quotas unless the official pricing page was read in full at the date recorded on the page. When this is not possible, the page says so.
Trademarks are used referentially. The site claims no partnership, endorsement, affiliation, or reseller relationship with any vendor it covers unless a current, explicit agreement exists and is disclosed.

How pages get to publication

Each page progresses through a documented status track maintained in content/content-status.json:

1. candidate — a slug is listed in the seed dataset but no page exists yet. 2. drafted — a page exists and is locally readable. 3. qa_passed — the page has been walked through the applicable section of qa/adsense-seo-quality-gate.md and every checkbox is satisfied. 4. ready_for_publish — the page is in the freshness window and the pre-publish checklist has been signed off. 5. published — the page is live at its public URL.

A page is never advanced ahead of its evidence. If a fact cannot be verified against an official source, the page either says so explicitly or omits the claim.

Update log

2026-05-22 — Initial pre-publication draft created. Placeholders remain in place; no domain or contact details have been finalized.