LLL · review-gated contract

Verification and Constraints

● READY FOR REVIEW

A complete, coherent review package for Rita, with useful choices for Matt and proof behind every client-facing URL.

Owner: Matt · executor not launched Run: KL-REVIEW-20260925-v2 Updated: 2026-09-25 · Prepare

Goal and finish line

End goal: Make every promised client-review asset inspectable at a verified public URL, with distinct alternatives and one unsent email guiding Rita through the package.

Done means: Matt can compare recommended and alternate ads, two landing routes, the email journey, fourteen daily units and the testimonial kit in one consistent website-derived design. Every page and link works in a bare unauthenticated URL, its key visual states have independent screenshot evidence, and Rita can see exactly which terms still need her decision.

0 / 28criteria verified

7deliverables

3+terms and operating details to confirm

Rules that stay fixed

Use the current site and Rita’s direct corrections as source authority. Keep the first pilot audience broad and the participant journey a clickable demonstration through payment confirmation, not a live service.
No client send, real upload, payment, ad activation, invented clinical outcome or binding commitment. Treat 24,900 Ft, 12 seats and three-working-day feedback as proposals until Rita confirms them. No invented launch date or countdown.
Host new review surfaces as bare unauthenticated URLs, without an added security layer. Keep private source data off public pages, preserve existing comments and unrelated work, and keep shared paid API spend within USD 15.

Proof standard: Each checkbox needs the stated artifact or URL readback. For every GUI check, attach an actual post-change screenshot, an independent blind gpt-6-luna high description of only what it shows, and a parent comparison to the criterion. Missing evidence stays unchecked.

End goal system

Two inputs → five outputs → Rita review

The two gray inputs jointly feed the green review outputs. The five outputs are parallel parts of one package, not a causal sequence. The route diagram below shows the review and execution order.

Input 1 · confirmed or open

Matt’s settled decisions

plans/rita-review-package-20260924/decisions.md

Why: Keep the two entry routes, full A/B alternatives and latest scope corrections intact.

Input 2 · constraint or source

Rita’s sources + website

plans/rita-review-package-20260924/landing-source-inventory.md

Why: Ground professional content, wording and visual identity in the eight completed audits and original material.

Deliverable 1 · main review item

Ads + landing journeys

Public review URL pending verification

Why: Let Rita experience both ways into the offer and Matt compare the complete A/B approaches.

Deliverable 2 · evidence or artifact

Email journey

Public review URL pending verification

Why: Show useful content and each transition from video feedback to enrollment.

Deliverable 3 · course

14-day material library

Public review URL pending verification

Why: Make every daily unit and original supporting document inspectable.

Deliverable 4 · testimonials

Testimonial kit

Public review URL pending verification

Why: Prepare six invitations and a practical interview and consent guide.

Deliverable 5 · review

Verified review index

Public review URL pending verification

Why: Gather useful variants, recommendations and tested review URLs in one place.

Output · decision or next action

Matt selects → Rita reviews

Why: The unsent client email explains every URL and the decisions Rita needs to make.

Next: Matt chooses variants, then sends Rita the reviewed email separately.

Route to the end state

Explicit review gate before execution

A review of this contract comes first. A later explicit Launch starts one executor. Publication and evidence come before Matt selects what Rita sees.

1 · review

Review this contract

Why: Matt checks scope and may annotate.

2 · gate

Explicit Launch

Why: Only a later Launch creates one executor.

3 · build

Complete assets

Why: Reuse drafts and sources; resolve gaps.

4 · prove

Publish + inspect

Why: Verify URLs, paths, copy and screenshots.

5 · handoff

Matt selects

Why: Compare A/B choices before Rita sees links.

6 · output

Rita reviews

Why: Unsent email tells her what to assess.

Parent ↔ child messages

Newest message first

PARENT · PREPARE25 September 2026

This board defines the verifiable client package. Matt reviews it before any LLL executor launch. Prior drafts and source audits are reusable, but no client-facing page is accepted by this revision.

Parent read setSmall files first

Executor read order after Launch: current decisions.md, landing-source-inventory.md, prior wrap-up, source audits and original material. Matt’s newest correction outranks an old draft; Rita’s direct comments and source documents outrank inferred copy. Reuse the partial local files only after checking them. Keep raw transcripts and private source material off public pages.

Source hierarchy: Matt’s latest corrections → Rita’s direct comments and meeting decisions → original course material → older drafts.

Future completion review requires a clean session gist, deliverable list and main-prompt list. None of these is a substitute for inspecting the actual assets.

Exact private source set: Project root /Users/agency/Documents/Agty/KomplexLogopédia. Read plans/rita-review-package-20260924/decisions.md, landing-source-inventory.md, outputs/rita-review-20260924/wrap-up.md, the existing outputs/rita-review-20260924/drafts/ and site/, then the source items cited by the inventory. Existing drafts are starting points, not verified deliverables. Keep personal data, raw transcripts, credentials and unpublished originals outside public asset roots.

Ideas to considerOptional

Later acquisition test

After the first pilot, compare advertising both routes with advertising only the video/list-building entry and selling cohorts by email.

Confirm operating capacity

Propose a practical free-video intake cap. Rita confirms workload before any three-working-day promise goes live.

Keep review simple

Recommend the strongest variants inside the shared index. Leave alternatives accessible without overwhelming the client email.

Work and proof

One card per deliverable

Create a review index for selection

Give Matt and Rita one simple route through the package, with meaningful alternatives and a single visual identity.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/client-index.html · public client index URL pending verified publication. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When the public client index opens at 390px and 1440px, then both entry routes and the recommended A/B choices appear in the hosted review index; attach both screenshots and the blind descriptions/comparisons.
When the current website and the review index are inspected side by side, then typography, colors, logo, imagery and core components map consistently across every hosted page; record source samples and screenshots.
When Matt selects A and B for each substantive asset in the review index, then two genuinely different complete versions open at named URLs or sections; record a variant inventory and comparison.
When the review index is opened in the in-app browser and its public files are searched, then all client routes open without login and no raw transcript, private prompt or credential appears; save URL readback and browser screenshots.

Protect

× One coherent brand system, not unrelated A/B designs.
× Use a small useful set of review URLs. Keep private operational notes out of public asset roots.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Build acquisition journeys to explain both offers

Deliver full free-video and direct-challenge paths from ad to simulated enrollment.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/campaign.html · public acquisition URL pending verified publication. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When A and B ad/creative combinations for both Facebook routes are opened, then a recognizable speech-related everyday situation, the reader’s goal and a concise purposeful CTA lead to matching landing promises in the hosted acquisition page; attach copy audit and screenshots. Reject unexplained speech-making setups and stiff “professional starting point” phrasing.
When A and B are selected for free-video and challenge offers, then two complete, distinct landing versions per offer appear in the hosted acquisition page; attach desktop/mobile screenshots and the source-copy map.
When A and B are clicked through on both acquisition routes, then video ad → free-video landing → demo receipt → sample feedback → consented email → challenge and direct challenge ad → challenge landing → simulated checkout → enrollment each reach the expected success, error, full and expired states in the hosted acquisition page; record separate route-by-route interaction results and screenshots without real upload or payment.
When proposed price, capacity, feedback timing and an unset date are inspected in the hosted landings, then proposals and opt-in are labelled accurately and no fake countdown or confirmed start appears; attach inspected copy and screenshots.

Protect

× Ads lead with a relevant situation, the reader’s goal and a concise CTA. Say short video and personal feedback; leave the exact minute limit and written-format detail to the landing. Avoid unexplained speech-making setups and stiff “professional starting point” phrasing.
× 24,900 Ft, 12 places and three working days are Matt’s proposals for Rita. No confirmed date or visitor-reset scarcity. No logged-in platform.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Write emails to guide the next step

Make the complete communication sequence useful, natural and ready for Rita’s content review.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/email-journey.html · public email journey URL pending verified publication. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When each journey stage is selected in the hosted email page, then two full Hungarian versions appear for receipt, sample feedback, nurture, weekly update, cohort promotion, enrollment and next steps; save stage-by-stage copy inventory and screenshots.
When a content-led nurture email is read from subject to CTA, then one useful speech-related observation precedes a natural challenge invitation in both versions; attach independent Hungarian copy review and exact text locations.
When timing, consent and buyer state are checked against the private journey map, then separate opt-in, weekly cadence, cohort-promotion replacement and buyer suppression appear with exact triggers; attach the map and scenario checks.
When every email CTA is followed in sequence, then it reaches the intended verified review page and the promise remains consistent with the previous stage; record link results, copy audit and screenshots of the hosted email journey.

Protect

× No fabricated personal assessment. Feedback remains Rita’s work, with examples explicitly labelled.
× No forced sentimentality, hype, em dashes or semicolons. No actual sending or invented deadline.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Assemble daily materials to support practice

Preserve Rita’s six modules in fourteen daily units and a complete downloadable review library.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/course-library.html · public course library URL pending verified publication. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When days 1 through 14 are opened in the course library, then each day shows a complete usable unit in a coherent order across all six sourced modules; attach day inventory, source map and actual page screenshots.
When original exercises and cautions are compared with the library and source downloads, then professional content is preserved and each download resolves; attach a source-to-day comparison and URL results.
When editorial versions A and B and reminder alternatives are selected, then both full presentations of the same canonical fourteen-day material and a closing-consultation guide appear in the hosted library; attach comparison and screenshots.
When recording and delivery instructions are read, then optional baseline, participant-held practice, at most one closing clip and proposed Messenger mechanics are clearly distinguished, without daily Rita feedback; attach copy excerpts and screenshots.

Protect

× Do not replace long-form source content with generic short summaries or invent clinical exercises to create variants.
× No daily Rita feedback, clinical guarantees or claimed live Messenger/login integration. No unsupported daily time commitment.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Prepare testimonials to collect useful feedback

Give Rita a complete invitation and interview kit and account for earlier client promises.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/testimonial-kit.html · public testimonial URL pending verified publication. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When the testimonial kit is opened and its invitation selector is used, then all six complete invitations and two coherent presentations appear in the hosted page; attach copy inventory and screenshots.
When interview and consent sections are inspected, then neutral questions and separate participation, recording and public-use decisions appear in the hosted kit; attach copy audit and screenshots.
When every booking CTA is activated in read-only mode, then the verified service-specific Reservio destination opens without creating a booking; record exact URL readback and browser screenshot.
When older reporting or therapy-booking promises are compared with the kit, then verified relevant links or precise unresolved gaps appear in its secondary review section; attach source mapping and screenshot.

Protect

× No invented testimonials, performance report data or unverified appointment duration in invitation copy.
× Reuse finished relevant work. Do not change live booking or create a new reporting project merely to fill the package.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Verify hosted pages to prove readiness

Prove that the assembled package works and reads correctly on the actual review host.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/site/verification.html · private QA evidence plus public URL checks. Resolve exact cited sources through the private landing-source inventory and record the final public URL after deployment.

Verify

When the complete public journey is read by an independent Hungarian reviewer, then all material copy contradictions are resolved in the hosted pages or logged as open; attach the reviewer report and corrected locations.
When navigation, A/B controls, demo states and downloads are tested by mouse and keyboard, then the expected route or state appears without real submission, purchase or enrollment; attach interaction log and screenshots.
When every visible section and meaningful state is captured after deployment at desktop and 390px width, then two independent Luna xhigh reviewers inspect the complete section/segment inventory and an independent blind gpt-6-luna high reviewer describes each GUI criterion screenshot; attach actual PNGs, reports and parent comparisons.
When any visual or functional defect is repaired and the stable URLs are reopened in a fresh unauthenticated client, then a fresh complete deployed section/state/viewport QA round passes, review comments persist and an older host page still opens; attach new PNGs, independent reports and URL readbacks. Leave unavailable checks unverified.

Protect

× Do not substitute another model and call it Luna. Preserve evidence of unverified checks.
× One publication owner preserves the shared host and stable comment identity. No force deployment that drops unrelated pages.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Draft Rita’s email to explain each link

Make the morning selection and client review easy without sending anything prematurely.

READY0 / 4

Show criteria, constraints, and evidence

Target and source: outputs/rita-review-20260924/drafts/email-draft-guided.md and email-draft-concise.md · private, unsent Rita drafts and morning overview. Resolve exact cited sources through the private landing-source inventory and record the verified public URLs referenced by the private drafts.

Verify

When each of the two unsent Rita email drafts is read, then every named client URL opens the matching final asset and each paragraph tells Rita what to review; attach drafts and URL readbacks.
When Matt opens the private morning overview, then recommended variants, reasons, decisions for Rita and exact launch/send boundaries appear before alternatives; attach the overview file and inspection receipt.
When the final asset inventory is compared with the overview and executor handoff, then every verified, proposed, untested and open item is classified accurately; attach inventory diff and exact resume path.
When the final Markdown voice transcript is reviewed twice by independent Luna Max critics and rendered, then a coherent Charon guide at 1.25× explains the big picture, morning review order and decisions in a playable local file; attach reviews, transcript and audio decode result.

Protect

× Do not claim drafts are an activated campaign or that Matt’s proposals have Rita’s approval.
× No automatic email, Telegram or other external sends. One combined final guide, no per-child audio.

Evidence slot: Pending. Executor adds the named artifact, exact URL or command output and parent comparison before checking a box.

Screenshot slot: Pending for every GUI check. Attach actual post-change PNG, independent blind gpt-6-luna high description and comparison. Keep unchecked until all three exist.

Open decisions and blockers

Only unresolved items

REVIEW GATE · no executorMatt reviews this board

Existing drafts remain partial. A later explicit Launch is required by LLL before one executor starts; the current request only prepares this board.

PROPOSALS FOR RITATerms are not approved

24,900 Ft pilot price, 12 paid challenge places and a short written video response within three working days are Matt’s suggested terms. Rita still needs to confirm price, paid cap, feedback workload, closing consultation, and whether Messenger delivery is feasible. Separately decide a sustainable weekly free-video intake cap and uploaded-video handling. Rita’s direct confirmation is the evidence for these operating details.

OPEN DATE AND CAMPAIGN CLOCKNo simulated scarcity

The start and close dates are unset. The review can show a labelled example countdown, but a live fixed deadline must wait for a real approved date. A future list-only campaign test remains an idea, not launch scope.

Recovery state

Only inspect on stall or completion

0consecutive stalls

NOT LAUNCHEDrecovery state

Local draftsresume point

Resume point: Prior local drafts and source audits remain in this project. They are inputs, not accepted deliverables. This review round starts at 0/28 verified.