Changelog

What shipped, when.

A public log of what we've added, expanded, or corrected on the JustineAI™ marketing site. The site evolves alongside the platform; the changelog makes the cadence visible.

Platform-side release notes — changes to the product rather than to this site — are published per edition: PI and WC. Security, access-control and confidential-data handling changes are not itemised on either page; your engagement carries that detail in full.

  1. September 13, 2026

    The refusals are now the pitch — and a gate that fails if one is ever removed

    • Three retractions today all pointed the same way: the honest sentence outsold the claim it replaced. When the Workers’ Comp filing-set claim came down and was replaced with “these are drafts, not the official state PDF forms — Justine fills no fillable form, and any field the record does not establish is left as a bracketed placeholder rather than invented”, the truth read better than the overstatement. So the refusals are now a band of their own, cross-edition, on the home page and on both edition pages.
    • It leads with the rating chain. The permanent-disability calculation flags its own output as unverified and labels an unverified adjustment table on screen, beside the number. In a practice where a wrong permanent-disability percentage is money out of a client’s settlement, an engine that names which of its own inputs it could not stand behind is worth more than one that always has an answer — and nothing else in this product is harder for a competitor to copy, because copying it means giving up the confident number.
    • Seven more follow, each in the product’s own words: Benson apportionment declines the comparison where the record carries only one industrial injury or a finding it cannot attribute; the deposition parser refuses a transcript whose page-and-line numbering it cannot read, names the line that failed, and stores nothing, because a guessed line number shifts every citation after it; an interrupted reasoning run reports itself interrupted and keeps the partial work rather than rendering as a finished answer; OCR returns an explicit not-configured result instead of faking a read; the approvals inbox keeps what Justine declined to do as a titled section and distinguishes “nothing is waiting” from “we could not ask”; outbound send is off unless a firm turns it on, so approving a draft sends nothing; and the statute-of-limitations table flags every entry that is not attorney-verified — fifty-one jurisdictions covered, four verified, all four Californian.
    • Every one of those is now a probe, not a promise. A new registry fetches the live PI and WC applications on each pull request and fails the build if a refusal’s own words have disappeared from the product. This is the mirror of the not-live registry and it guards the direction nobody watches: a refusal arm is an early return, it looks like a dead branch to anyone reading it cold, and deleting one produces no failing test and nothing resembling a marketing event — while this page goes on advertising a guard-rail that is gone. The needle sets differ per edition on purpose, and the negative control was run to watch all eight claims fail and recover before any of it was published.
    • The band closes on what is armed, because a page of refusals with nothing behind it is its own false impression: 391,106 agent events across the production deployment in the seven days to 13 September, over 341 tasks; 178,167 rows in a hash chain where each carries the HMAC of the row before it, written through one serialised writer per tenant and verified nightly; 3,301 per-case access grants; and fail-closed tenancy actually switched on rather than staged behind a flag. Those measure the machinery, not adoption — Justine is in market with no live customers yet, and the figure the catalogue carries as “0 of 9 firms have enabled outbound send” is published here without the 9, because a denominator of firms is a customer count whichever way it is pointed.
    • The pricing matrix stopped selling a ten-million-token context window. That row promised the full case file in one inference; measured against the backend this morning, documents are split into 3,500-character windows, retrieval returns eight of them, each truncated to 2,000 characters, and the largest input budget anywhere in the product is 40,000 characters. The gap is about a thousandfold and it sat on the row an evaluating attorney reads hardest. It is replaced by something defensible line by line rather than reworded. The same claim is still published elsewhere on this site and is being worked through separately — the long-context slot is an architecture claim shared across four sibling sites, and pulling it from one of them alone would leave the four disagreeing.
    • The PI release notes were three entries against fifty-six pull requests merged since 20 August. Thirteen more are published, written from the merge history rather than from memory, and they are mostly corrections: a negotiation figure that could display at a hundred times its value, a 33.33% contingency fee the product had invented and then printed on a client’s disbursement, a deadline that could save without its due date, report filters that filtered nothing under a heading that named them, and an audit export that truncated at five hundred rows while reporting that there was no more. The withdrawn Bates entry stays withdrawn.
  2. September 13, 2026

    Eight places where this site disagreed with itself

    • The pricing page opened by saying we do not believe in vendors who hide their pricing, and then published no prices — while /pi had been publishing $129 and $349 per seat the whole time, under tier names that did not appear on the pricing page at all. The seat rates are on /pricing now, read from the same constant /pi renders so the two cannot drift, and the page explains the thing that was actually confusing: a seat is priced by who sits in it, a tier is the procurement shape a firm buys under, and the tier does not change what Justine can do.
    • Three different counts of the same thing were published as facts. /pricing said "All 21 PI capabilities", /pi said 34, /pi/capabilities said 77+. None of them was lying on purpose — they count different granularities — but a prospect reading two of the three learns only that we do not know. /pricing no longer publishes a number at all (the catalogue is not tier-gated, which is the only claim that mattered), and /pi now says its number is the headline set and links to the longer catalogue.
    • Billing was described as running on Stripe and Square across four surfaces, including the sub-processor list on the compliance page. There is no Square path. One processor, named accurately, and the compliance page no longer names a payment processor that receives nothing.
    • The home page said MindHYVE serves institutions across four regulated verticals, forty-four lines below the portfolio statement — on the same page, from the shared canon — that says five. The corporate-structure pages listed four operating LLCs and omitted Eve-Finance, LLC, which operates EliAI and is named as the operator on eligrid.ai. Five in all five places now. The 2024 entry on the founding timeline still says four, because in 2024 it was four.
    • Two entirely different nine-item pipelines were both called "the 9-step Curation Layer". One is nine modules that clean the record on the way IN — OCR Janitor, Provider Canonicalizer, code validator, and so on — before the reasoner sees anything. The other is nine attorney-attestable steps that build the demand package on the way OUT. They share a count and nothing else. The first keeps the name Curation Layer and is now described by its module count; the second is the demand-package build, and the page that explains it says plainly that it is not the other one.
    • The ROI calculator cited "docs/STRATEGY.md §3 Tier 2" as its source. That is a path inside a private repository — provenance nobody outside the company can follow, which is the same as no provenance. It now says what it actually is: our own internal modelling, not a published study and not measured at a customer, because there are no live customers yet and we will not imply otherwise.
    • The team page dated its headcount "(May 2026)" in prose beside a number read from a constant — and the constant had carried the correct month, August 2026, along with a comment explaining exactly this failure, since the day it was written. The date is now read from the same place as the number.
    • The footer flattened every forthcoming edition to "· soon". The registry distinguishes Mass Tort, which is coming next, from Medical Malpractice, Insurance Bad Faith and Employment Class, which are on the roadmap — a distinction a firm deciding whether to wait actually cares about. The footer reads the registry now instead of restating it.
    • Bates stamping came off the export claims. The stamping routine exists in the code and no production path calls it, so no stamped PDF is ever produced.
  3. September 13, 2026

    The signing retraction finished — nine more surfaces, and a probe so it cannot come back

    • The e-signature retraction earlier today corrected two FAQ answers and stopped. It had been published as shipped on ten surfaces, and the other eight went out unchanged — /for-attorneys listed it as a practice-operations capability, the pricing matrix included it in all three tiers, /pi/capabilities and the PI grid counted it twice, and the compliance page named DocuSign as a sub-processor. A retraction that lands on two of ten surfaces is not a retraction; it is a footnote under a claim that is still selling. All ten now say the same thing.
    • DocuSign came down with it, which was not part of the original finding. Every signing sentence on this site paired the two — native e-signature as the default, DocuSign as the bring-your-own alternative — so retracting one and leaving the other would have left the page implying a firm could still get a document signed. It cannot. Measured against 2.1 MB of the served PI application and 1.6 MB of the served WC application: neither contains a signing route, the ESIGN or UETA labels, the Certificate of Completion, or the string DocuSign even once — while Dropbox, OneDrive, Outlook and the Microsoft sign-in route are all plainly present in the same bytes. The integrations surface ships; signing is not on it.
    • Both claims are now defended by a probe rather than by anyone remembering. They are entries in the not-live registry, which fetches the live product on every pull request and fails the build if a needle it expects to be absent ever appears. The needles differ between the two editions on purpose: ESIGN and UETA are absent from PI and are good needles there, but each appears once in the WC bundle, so using them on WC would have failed against a correct site, been silenced, and taken the whole claim with it.
    • The one figure this moved is the practice-operations count on /for-attorneys, from twenty-six to twenty-four, and the page now says why the number went down rather than letting it drift quietly. /pi/capabilities drops from 77 to 75 for the same reason. A count that only ever rises is not a count.
    • What is still true, and is what the copy now says instead: engagement letters, releases, settlement agreements, demand letters and the WC settlement papers are drafted and exported as Word or PDF on the firm’s own letterhead, rendered server-side, for signature out of band. The envelope, its hash chain and the certificate are built and waiting on a mail provider and a signing page. That is a roadmap item close to shipping, and it is written here as one.
    • Bates stamping came off the export line in the same pass. The stamping routine exists and nothing in production calls it, so no stamped PDF is ever produced — the same shape of defect as the signing claim, found in the same audit, and removed rather than reworded.
  4. September 13, 2026

    Nine published claims withdrawn, and the two gates that let them stand

    • A forensic audit read the site against the product repositories and found the drift running the wrong way — claims ahead of the code rather than behind it. Nine are withdrawn here. Each is a narrowing, not a rewrite: the sentence comes down to what the code does and nothing is put in its place, so each can be restored the day its feature ships.
    • The pricing page sold multi-factor authentication. It does not exist — there is not one source file in the backend that mentions MFA or TOTP — and our own Trust Center already said so, in the correct words, on a different page. Two published artifacts disagreed about a security control and the false one was the more confident. The pricing page now says what the Trust Center says.
    • Single sign-on is OIDC, including Microsoft Entra ID, and five surfaces were selling SAML alongside it. SAML is not implemented anywhere: it appears in the product as a catalogue label and a help-search keyword. The pricing matrix now states that it is not supported rather than listing it as an enterprise inclusion.
    • Workers’ Comp did not “assemble the state’s filing set” and never has. There is no form-filling machinery and not one form asset in the product. The Application for Adjudication and the Declaration of Readiness are drafted as prose on the firm’s own letterhead, with every field the record does not establish left in brackets rather than invented; the DWC-1 is not drafted at all — Justine recognises it on ingest and dockets the §5402(b) clock from it. The copy says drafts now, and names the DWC-1 for what it is.
    • The rating claim was right about California and wrong about the other two states. The California chain is verified end to end and a standard post-2013 request is returned on an official basis — that part stands, and it is worth more than the sentence that overstated it. Georgia and Texas do not rate: they take the evaluator’s impairment rating as an input and compute benefits from it. Four surfaces said all three states were fully rated.
    • The California rule corpus was published as 399 in four places and 400 in one, six weeks after the correction from “~400” was made. The coverage table was typing its own copy of a number that already had a home. It now derives from that home, which removes the second source rather than resynchronising it.
    • §2 of the Terms of Service said Workers’ Comp was on the roadmap and not generally available, on a page whose own navigation said it was in market. The registry says in market, so the contract was the stale document. It now derives the roster from the registry instead of restating it, and so does the Founding Firm programme, which was still promising WC beta seats “when WC opens”.
    • “IOLTA-grade” came off the trust-accounting copy. The reconciliation is real and it is three-way, but its third leg is the adjusted bank balance, not the sum of the per-matter sub-ledgers — which is what the phrase means to a bar auditor — and the bank balance is typed in by a person because there is no bank feed. One WC line claimed the sub-ledger leg outright. A compliance adjective is not ours to attach.
    • In-app e-signature was published as live. The envelope, the audit chain and the certificate are built; no mail provider is wired and there is no signing page, so no signer can finish. It is on the roadmap now, in the words this site already uses for work that is finished and unreachable — we count it as nothing until you can reach it.
    • A real insurance carrier was named on /technology/agents beside an unsourced statistic about how it settles. The same worked example is anonymised everywhere else on the site; it is anonymised there now too.
    • Two gates were repaired, because eight of these nine were things a gate was supposed to catch. The edition-status check could not see the Terms of Service at all — the file sat outside the paths it scanned — and would not have caught the sentence even inside them, because it hunts a capitalised badge literal and the Terms said “on the roadmap” in ordinary lower-case prose. Widening the paths alone would have been theatre. It now also reads availability prose against the registry and fails when one sentence says an in-market edition is unavailable; that rule was run against the original sentence before it was trusted.
    • The claim-registry scan was pinned to a version of the registry four releases old, which is the failure the pin’s own comment predicted in writing: it enforces the rules of the day it was set and stays green while the registry moves. It is current now.
  5. September 9, 2026

    A fifth sibling, a not-live claim that had stopped being true, and where you came from

    • EliAI joins the family. The portfolio statement, the footer, the landing family band and /about/mindhyve-family now name five Agentic Operating Systems rather than four, and link to eligrid.ai — which had been linking to all four siblings while none of them linked back. The wording is copied verbatim from the shared canon rather than typed here, and the same change landed on the Arthur, Chiron and Theo sites in the same pass, because nothing in the estate detects one site stating a different portfolio from the others.
    • Push notifications came off the “what isn’t live” list for Workers’ Comp — 19 entries down to 18. The site was telling attorneys there was no push channel; the shipped WC bundle has a real subscription path and the setting is rendered unconditionally in Notifications, next to the deadline reminders. The claim was checked against the live product before it was deleted, because the other explanation for a gate like that going red is that the gate is wrong, and deleting the claim to make it pass is how a not-live list quietly turns into fiction.
    • Signup links now carry first-touch attribution to the product portals. When a firm meets us through a referral or a campaign, disappears for a week and comes back by typing our name into a search box, the channel that actually introduced us is the one recorded — not the search that closed it. No visitor id, no fingerprint, no cookie, no third-party pixel is minted: the fields carried across are the campaign parameters an advertiser put in the link, the referring site, and when you first arrived.
  6. September 5, 2026

    Six WC surfaces the site had not caught up with — and two not-live claims that had gone stale

    • The WC capability catalog was re-audited against the product repositories’ master and, this time, against what the live surfaces actually report — the WC app and its API each publish the commit they are serving, and both matched the commit that was read. “Reachable by an attorney today” is the bar this page claims to hold itself to, so it is now a checked statement rather than an assumption about deployment.
    • Six surfaces added: trust accounting with a three-way reconciliation and a per-matter sub-ledger; a matter-scoped portal the injured worker signs in to; search across the firm and across a single matter; deposition transcripts addressable by page and line; importing an existing book of matters from a CSV export; and records requests with document blockers, alongside a notification tier that cannot be muted.
    • Every one of them is published with its limit attached, in the entry rather than in a footnote: search never reads all four corpora in one query and cannot narrow document search to a matter; deposition ingest takes roughly forty-page excerpts of plain text, not full volumes or PDFs; case import creates matters one at a time and is not a transaction, so a file that fails partway leaves the earlier rows already created; and trust accounting is a records surface, not an audit — no bank is read, and a reconciliation forced over an unexplained difference is recorded as forced.
    • Two entries in the “what isn’t live” list had quietly become false, which is the direction of drift nobody looks for. The site was still telling attorneys that the injured worker gets a letter and not a login, and still counting page-and-line deposition citation as finished code with no screen. Both had shipped. A not-live list that is never re-read stops being honesty and becomes an understatement of the product.
    • The client portal is published as what it is — live at the client’s end, unfinished at the firm’s. An invited client can sign in; issuing that invitation and releasing a document are API operations with no screen in the workspace yet, and the product does not email the invitation. That half is named in the not-live list rather than rounded up into the capability.
    • No published figure moved. The headline counts are floors — 240+ capabilities, 135+ features, 90+ AI and agentic — and new surfaces leave a floor true, so nothing was restated. An exact re-census was not attempted here and no count was inferred from six additions.
    • The catalog’s group count is now derived from the groups themselves. It had been the word “nine”, typed into the headline beside an item total that was computed — one fact with two sources, in the sentence that introduces the page.
  7. August 29, 2026

    Capability re-census · a WC catalog · three figures corrected down

    • The capability census was re-derived from the live PI and WC codebases, against origin/master in all three product repositories rather than a working checkout — two of the three were sitting on branches behind master, and every count taken from them would have been an undercount. The previous census was taken on 2026-08-16, before the largest build window the product has had.
    • New /wc/capabilities — the full Workers’ Comp catalog, in nine groups. WC already had an overview, a how-it-works and release notes, and no page that inventoried what had been built behind them; PI had one. The page counts only what an attorney can reach — a route they can click, or something they can ask Justine for — and names, at the foot, the finished and tested work that has no screen yet and is therefore counted as nothing.
    • Headline counts moved up: 240+ customer-facing capabilities (was 230+), 135+ user-facing features, 90+ AI/agentic capabilities (was 80+), and 10 background automations (was 7). A fourteenth capability group was added for the agent fleet and its approval rails — nine named agents, one authorization step that lets an agent read, compute, draft, save and notify inside the firm but only ever QUEUE anything that sends outward, files outward or touches money.
    • Three published figures were wrong in our own favour and have been corrected down or scoped, rather than left standing. Statute-of-limitations coverage was published as 52 jurisdictions; the table holds 51 (50 states and DC) and always did. Compliance and security controls were published as 24; the enumerable inventory lists 14, and 24 could not be reproduced from any artifact, so the number published is now 14. The California rule corpus is stated as 399 rather than “~400”, and the statute-of-limitations table is no longer described in a way that implies it shares the rule corpus’s author-to-reviewer verification, because it does not.
    • One figure was corrected in the other direction, because it undersold the product: roughly four in five of Justine’s AI capabilities are deterministic legal and actuarial engines — statute math, rule lookup, date computation — not model-composed output. The site had been saying “roughly half”. 27 of the 35 reasoner-callable tools never reach a foundation model.
    • Correction to the entry of July 11, 2026 below: it described the public intake portal as live. The intake portal’s server side is real, but the public page and the endpoint it calls do not currently meet, so a firm’s generated portal link does not resolve end to end. The claim has been withdrawn from that entry and is not made anywhere else on the site until it does.
    • WC and PI release notes brought current for the August window: the approval inbox and the record of what Justine declined; the §4610 utilization-review clocks; Kite reported alongside the Combined Values Chart with neither selected; the substantial-evidence screen on apportionment; drafts that can be opened, edited, versioned and exported on the firm’s own letterhead; Spanish letters verified to be in Spanish; and — on PI — the fix for an interrupted reasoning stream that used to report itself as a completed answer.
  8. August 27, 2026

    Disclosure scoping synced, and a corpus gate that pinned the wrong number

    • The rule governing where a foundation-model name may appear is now synced from the shared canon rather than kept as a local override, so it cannot drift per site. The customer-facing label is the Eve engine; the model underneath is not named on customer surfaces.
    • The IPSC works count is 123, not 128 — and the gate meant to protect that figure had been pinning the wrong one, so it was enforcing the error rather than catching it. A check that agrees with a mistake is worse than no check, because it converts a typo into a defended position.
  9. August 21, 2026

    Six-month headcount growth published

    • The team figure now shows growth over six months rather than a single number, and the published band was widened to hold it. A point figure invites the question of when it was true; a band with a trend answers it.
  10. August 20, 2026

    Claim gates, per-edition release notes, and five figure corrections

    • Edition status and published figures are now gated in CI. Both had been hand-maintained across many pages, which is the arrangement that produced the five corrections below.
    • Per-edition release notes were added, so what shipped in an edition is stated where a buyer of that edition looks.
    • SOC 2 copy asserted more than the claims registry supports: it described the report as covering our controls, in the present tense, and referred to the attestation as already held. The registry lists it as pending. The copy is now future tense throughout. Claiming an attestation you do not yet hold is the single most expensive sentence to get wrong on a legal-technology site, and the gate that enforces this refuses the phrasing outright rather than trusting a reviewer to catch it — including, correctly, in this entry’s first draft.
    • Headcount is 212, propagated from the canonical source rather than restated per page. The publicly-listed count was a literal that disagreed with the roster it was supposed to describe. The IPSC card quoted a corpus two versions old, and a gate now pins it.
    • The portfolio statement was synced from the corrected ArthurAI copy.
  11. August 16–17, 2026

    Machine-readable corpus, all-edition capability showcase, model SKUs removed

    • A machine-readable capability corpus was published, so an AI crawler reads the same inventory a person does rather than a thinner summary of it.
    • /capabilities now shows every edition, census-exact, instead of describing a subset in prose.
    • Foundation-model SKUs were taken off the landing and Trust surfaces. Naming the model underneath is barred on customer surfaces, and the Trust Center is precisely where a reader is entitled to assume the copy has been checked.
    • Cross-product deployment language was softened to engagements, matching the instrument each claim actually rests on.
  12. July 28–31, 2026

    Lead attribution, consent model, and share cards

    • First-touch attribution was added so a lead can be tied to the campaign that produced it.
    • The consent banner had been covering the contact form on mobile — blocking lead capture outright on the surface the site exists to drive. Fixed, and analytics moved to a CCPA opt-out model.
    • Brand-forward social share cards are now generated per page, so a link shared anywhere unfurls as the page rather than as a generic site card.
    • An editorial rule was overflowing its container on mobile.
  13. July 19–27, 2026

    Workers’ compensation: multi-state grounding and a statute demonstration

    • The WC edition surfaces now show the multi-state grounding and a statute demonstration, so the claim that Justine reasons against jurisdiction-specific law is shown rather than asserted.
    • A promise of a self-serve trial start was removed. The trial is assisted, and saying otherwise sets an expectation the product does not meet at the moment a buyer is most attentive.
  14. July 11, 2026

    Workers’ Comp in market · capability reconciliation

    • Workers’ Comp is now in market for applicant-side California practice. New /wc product page in the burgundy WC identity: the PD Value Engine (WPI → FEC → occupation → age; strict-AMA vs. rebutted), apportionment reasoning (Escobedo · Almaraz-Guzman · Ogilvie), AME/PQME analysis, MMI/P&S, UR → IMR tracking, C&R-vs-Stips settlement with net-to-client, the WCMSA flag (analysis, not a dollar calculator), and bilingual EN/ES client letters + EN/ES voice. Rating and settlement figures are projections the attorney verifies.
    • Under-sold live features surfaced: client SMS with an attorney-alert number-privacy relay, document RAG (“ask over your documents”), and — on WC — bilingual English/Spanish client letters and voice. A new Client communications section on /pi and /wc.
    • Roadmap items are now clearly labeled as coming, never implied live: Filevine integration, a client case-portal, and phone-intake / telephony. The turnkey approved live integrations are Microsoft 365 and Dropbox; DocuSign is available as a bring-your-own connection (firms supply their own integration key to view their envelopes, not send). Native e-signature (ESIGN/UETA) is the default. (Corrected 2026-08-29: this entry originally described the public /intake portal as live. See the entry above. Corrected again 2026-09-13: the two sentences about DocuSign and native e-signature were wrong when written — neither is reachable in either edition, and both are now on the roadmap. See the entry at the top of this page.)
    • “Predictive analytics” softened to “outcome statistics” — heuristic firm and venue statistics, not a trained ML model.
    • Edition roster updated across the site: PI and WC (applicant-side, California) in market; MT, MM, IB, EL on the roadmap.
  15. June 19, 2026

    Bias architecture on the record, edition framing

    • New bias-architecture section on the Trust page, in the portfolio voice: you cannot train bias out of a model — you separate the reasoning from the knowledge from the jurisdiction, so the bias becomes something you can read, audit, and govern. We describe population-free reasoning plus runtime steering; we never claim the system is “unbiased.”
    • Six-edition roster framing equalized — PI in market; MT, WC, MM, IB, and EL on the roadmap, each with a status badge so the footer ships no dead links.
    • Shared portfolio canon refreshed: the cross-site “bias is architectural” sentence added, and the portfolio statement updated to “four editions in production — ten across the roadmap.”
    • Added this public changelog and wired it into the footer.
  16. June 18, 2026

    Single-typeface system

    • Typography unified to a single typeface — IBM Plex Sans across display, body, and italic — matching the MindHYVE Brand Bundle canon now shared across all five MindHYVE marketing sites.
  17. June 17, 2026

    Frontier-component abstraction

    • Third-party frontier model names abstracted on customer surfaces; Microsoft and Azure remain named as the platform, and customer copy uses the Eve-Legal F5/reasoner label.
  18. June 6–8, 2026

    Deploy + asset hardening

    • Unknown routes now return a real 404 status instead of a soft-404.
    • Replaced 404ing favicons with PNGs from the JustineAI™ medallion; corrected broken asset and essay links.
    • Contact emails routed to @mindhyve.ai role addresses.
  19. May 30, 2026

    Consent, team, legal, Eve-Genesis

    • Added a GDPR/CCPA consent banner; Microsoft Clarity is gated on consent.
    • Added a /team page and retired /about/leadership; the leadership block now points to the team roster.
    • Stripe activation commerce policies standardized under /legal/*.
    • Eve-Genesis page gained the cognitive-fingerprint and riddle-origin reveal.
  20. May 19, 2026

    Initial production deploy

    • justinegrid.ai launched as an editorial site: /pi (Personal Injury, in market), /technology (Eve-Legal F5/reasoner), /about, /trust, /contact, and /legal.
    • Six-edition roster — PI in market; WC, MM, MT, IB, and EL on the roadmap — plus the ROI calculator and the Founding Firm program.