Field log

What shipped,
and what hasn't.

Every capability’s state, counted from one release table; the dated releases; the open questions; and the original field log with its corrections beside it. Nothing planned is presented as shipped.

Shipped3
Beta2
Planned27
Research questions5
Capabilities tracked32
01 // Release board

Every capability.
One state each.

Grouped by state, then by the surface it runs on. Each row keeps its exact label and links to its page.

Board // States, counted32 tracked
Available
3 capabilities
Beta
1 capability
Private Beta
1 capability
In development
5 capabilities
Planned
22 capabilities
Coming soon
0 capabilities
Latest dated release
2026-08-17 — Founders became a request; every seat count came down
Scoring versions
Rubric v1.0 · pillar layer v1.0

Beta

2 capabilities · Beta + Private Beta

Runs today with a stated limit: not yet verified end to end, or open only to people Rank Sniper enrols.

Planned

27 capabilities · In development + Planned + Coming soon

Does not run yet. In development means work is under way; Planned means it is designed, not built.

02 // Dated releases

Shipped, with
the date on record.

The only dates on this page, each read from the release manifest.

  1. 2026-08-17Founders became a request; every seat count came down

    A signup is a no-obligation request with no payment field; seat counters were removed because nothing is reserved or sold here.

  2. 2026-08-16RankSniperBot documented, and robots.txt obeyed

    The scanner’s user-agent now points at a page listing exactly what it fetches, and the scanner reads robots.txt first. Before this release it did not.

  3. 2026-08-14Scan results kept out of the search index

    A scan result describes a store that did not ask to be listed, so it serves noindex and stays out of the sitemap.

  4. 2026-08-11Support tickets in HQ, answered from both ends

    Merchants open, read and answer tickets in HQ Merchant Command, in the same queue staff work.

  5. 2026-08-10Structured data on the English site

    A server-rendered JSON-LD graph cross-referenced by @id. Eligibility only: whether any rich result appears is Google’s decision.

  6. 2026-08-06The catalogue scan and evidence report

    The public scan of any Shopify storefront, scored against the 11 scored v1.0 checks: the first capability recorded as working end to end.

03 // Research

Open questions.
Not promises.

Stated as questions. None is scheduled, and none changes a score until it ships under a new version.

RSR-01

Does any AI vendor read llms.txt?

Knownllms.txt is a community proposal, not an adopted standard. Google states that Google Search ignores it and that publishing one neither helps nor harms visibility there. The crawler documentation OpenAI, Anthropic and Perplexity publish does not mention reading it.

OpenThe GEO pillar checks for the file today. Whether that check keeps its weight depends on evidence of use that no major vendor has documented.

RSR-02

Policy allows it. Does the agent get the page?

Knownrobots.txt states a policy, not a delivery: RFC 9309 calls it “not a substitute for valid content security measures”, and a firewall or CDN can still refuse an agent the file allows. When a declared request is hard-refused, our own crawler retries it with a browser profile and records which path served the scan (/bot).

OpenThe scan reads policy only. Testing what each agent actually receives belongs to post-fix verification (Planned); the question is how to measure it without misrepresenting who is asking.

RSR-03

Which agents belong in the penalty?

KnownThe scan checks 10 agents. Anthropic also documents Claude-SearchBot and Claude-User, and its current article does not list anthropic-ai, which the scan still checks. Perplexity says Perplexity-User generally ignores robots.txt. Google says Google-Extended does not affect inclusion in Search; Apple says Applebot-Extended does not crawl webpages.

OpenWhether to add agents, retire legacy ones, or score control tokens apart from crawlers. Any of it moves scores, so it would ship as a new rubric version, never silently inside v1.0.

RSR-04In development

One Rank Score instead of two layers

KnownThe composite (rubric v1.0) and the pillar layer (v1.0) come from one audit, read some of the same tests at different weights, and are reported side by side.

OpenWhich checks, which weights and what coverage statement one versioned contract should carry, so every surface reports a single number.

RSR-05

What would an honest benchmark need?

KnownOnly demo catalogues have been measured end to end, so no median or zero-touch rate is published.

OpenHow many real stores, sampled how and with whose consent, before an industry number says more than its sample.

Sources for the research questions
  1. OfficialOptimizing your website for generative AI features on Google SearchGoogle Search Central
  2. ExperimentalThe /llms.txt file (a proposal)llmstxt.org
  3. OfficialRFC 9309: Robots Exclusion ProtocolRFC Editor
  4. OfficialDoes Anthropic crawl data from the web, and how can site owners block the crawler?Anthropic
  5. OfficialPerplexity crawlersPerplexity
  6. OfficialGoogle’s common crawlers (Google-Extended)Google Search Central
  7. OfficialAbout ApplebotApple
04 // Field log

The original log,
corrected beside it.

Written from the repository record. Each entry carries the date it became true in a deployed lineage, its category, and the commit that carries it.

Shipped

  1. September 14, 2026 · App · 74ed362Feed endpoints and webhooks, built and reconciling in code

    The canonical record serves in four formats and reconciles against the database two independent ways. Product-update webhooks re-enrich on change, with a loop guard so our own write does not wake us forever.

  2. September 14, 2026 · App · 74ed362The write layer, and a contract enforced in code

    Enriched fields land in one dedicated metafield namespace. Every other mutation is refused at the API client before the request is issued — an allowlist, not a policy. Because the writes are confined to that namespace, removing it leaves merchant-authored fields exactly as they were; a merchant-facing revert control is planned and not built.

  3. August 22, 2026 · This site · 3a72091Overwatch, wired to the same sources the pages render

    The assistant answers from a knowledge file generated at build time out of the same pricing, pillar and rubric sources the pages use — so it cannot quote a price the site does not show. A gate fails the build if the two drift apart.

  4. August 22, 2026 · This site · 4c061efV3 Fire Control — the marketing site, rebuilt

    One design system across the marketing tree, with every number on the page read from the same source the product uses. Anything that could not be traced to a source was removed rather than estimated.

  5. August 22, 2026 · Commercial · 4c061efPriced by catalogue, never by pillar

    All four disciplines are in every tier. A tier decides how large a catalogue is read and how often — it never decides which pillars a merchant gets.

  6. August 22, 2026 · Measurement · f6881dcThe measurement audit became a standing gate

    Three checks were found measuring something other than what they claimed — one scored every store on earth 0% on a tenth of its SEO pillar. The audit that found them now runs on every rubric change, because we expect to find more.

  7. August 22, 2026 · Foundation · 4c061efCrawler policy, read by name

    robots.txt is parsed against 10 named AI agents rather than pattern-matched. A blocked agent is a capped penalty in the score — 6 points of the composite each, to a maximum of 30 — because that is what the rubric does.

  8. August 22, 2026 · Foundation · 4c061efThe scanner and the published rubric

    A public scanner that reads any Shopify storefront and scores it against 4 pillars, with every catalogue check and its weight published at /rubric rather than described. Rubric v1.0, versioned so a score always says which rules produced it.

In build

  1. September 15, 2026 · In buildBefore/after shot reports

    The write pipeline is real and the scoring is real. What does not exist yet is a store scanned, enriched and then re-scanned — so there is no measured "after". The section is built and deliberately renders nothing until there is one.

  2. September 15, 2026 · In buildMore attachments

    No attachment is orderable today. Each one is designed and tagged to the pillar it serves, and each is marked planned on the pricing surfaces until it is built.

  3. September 15, 2026 · In buildStorefront JSON-LD block

    The hosted feed works today. The theme extension that publishes the same record on your product pages needs one toggle in your theme editor — we do not request permission to edit themes, which is why it is your switch and not ours.

  4. September 15, 2026 · In buildFull French and Spanish

    Navigation, hero, footer, pricing and the home page sections are translated. The knowledge layer — the journal, the field guides, the lexicon and the crawler index — is still English only, so the site does not yet declare French or Spanish alternates to search engines.

Not claimed yet

  1. September 15, 2026 · Not claimedA published benchmark

    No zero-touch rate or industry median is published, because the only catalogues measured end to end so far are demo data. A number about sample data wearing this product’s name is worse than no number.

  2. September 15, 2026 · Not claimedAgent traffic figures

    Nothing here reports agent requests served or referred revenue. Those are not measured yet, so they are not shown — the hero counters and the command centre suppress themselves rather than estimate.

  3. September 15, 2026 · Not claimedA refund or trial offer

    There is none, deliberately. If one is ever introduced it will be written fresh, not revived from a retired draft.

05 // Engage

Test what shipped
on your own store.

Run the scan that shipped, then read the rules behind the number.

Every ruling behind these lines is dated and kept, including the ones that reversed an earlier decision. If a claim on this site ever looks unsupported, ask us for the entry.

Rank Sniper Changelog — Shipped, Beta, Planned and Research