The voice agent explains any task, answers questions grounded in this plan, and records a brain dump. Voice feedback is read-only and cannot approve or execute work. Allow the microphone when your browser asks.
The explicit Sonnet rollback keeps the existing capsule-only protocol. A lost connection retries the same Sonnet engine twice and reloads the complete persisted conversation. It never changes models during a conversation.
Sonnet 5 · no fallbackVoice answers are review evidence. Approval, sending, spending, and production changes stay in the originating chat. Voice trouble in this page? Open the standalone voice page.
Plan: /Users/agency/Documents/Agty/KomplexLogopédia/action-plans/2026-08-13-kl-14-day-challenge-review.md
J or K, to open it.1 delegate, 2 drop, 3 idea, 4 ask first, 0 clear, J and K move between tasks.!.The launch date stays open until the complete funnel and campaign are visible and approved. The participant model uses an optional opening selfie, practice recordings kept by the participant, at most one final 2 to 3-minute practice video, and a closing Google Meet. Exact cadence, offer, price, capacity, consent, storage, and campaign variant remain decisions.
Create one versioned requirements and decision ledger that reconciles the latest meeting, Slack files, Figma feedback, and current funnel assets. Separate approved behavior from agency proposals and superseded concepts.
The challenge is spread across transcripts, comments, PDFs, landing drafts, and older campaign concepts that contradict one another.
Builders may reintroduce repeated feedback, unsupported claims, unapproved prices, or an assumed launch date.
Every later asset can be checked against one evidence-bound source of truth, while Rita reviews only the decisions that remain open.
A dated ledger maps every confirmed rule, unresolved decision, rejected concept, source, and affected asset. 80/20. One decision matrix covers journey, workload, claims, data, campaign variables, and the launch gate.
flowchart LR A[Scattered evidence] --> B[One decision ledger] --> C[Safe build contract]
flowchart TD
S1[Read latest evidence] --> S2[Separate facts and proposals]
S2 --> S3[Map rules to assets]
S3 --> D{Every item sourced}
D -- yes --> DONE[Ledger verified]
D -- no --> S2Create a native Google Slides deck that presents the complete challenge funnel, the main viable variants, the evidence behind each choice, and the decisions required before choosing a launch date.
The funnel is fragmented across landing prototypes, PDFs, email ideas, ads, booking screens, and discussion notes.
The team keeps discussing isolated assets without seeing workload, dependencies, decision consequences, or the minimum preparation window.
One presentation enables a focused decision meeting and makes launch timing the consequence of an approved plan.
An editable deck covers the full journey, multiple variants, ownership, compliance gates, and the launch rule, with proposals labelled. 80/20. Seventeen decision-oriented slides with three core funnel variants and recommendation logic.
flowchart LR A[Fragments] --> B[Complete Slides story] --> C[Client decisions] --> D[Launch window]
flowchart TD
S1[Ground in ledger] --> S2[Storyboard journey and variants]
S2 --> S3[Build native Slides]
S3 --> S4[Run visual QA]
S4 --> D{Evidence and flow pass}
D -- yes --> DONE[Deck ready for review]
D -- no --> S2Update the challenge-content specification and drafts so all six modules, the landing page, and instructions consistently reflect Rita's latest feedback.
Several drafts still describe repeated submissions, broad questionnaires, conflicting progress framing, shortened teaching copy, or challenged wording.
Participants may receive the wrong workload and Rita may see a version that ignores her comments, delaying approval.
The content becomes consistent, simpler to deliver, and visibly faithful to Rita's professional model.
All six modules and the landing and delivery specification pass a comment checklist, preserve substantive teaching text, and surface remaining choices. 80/20. A corrected journey and redline checklist.
flowchart LR A[Conflicting drafts] --> B[Comment aligned content] --> C[Coherent participant journey]
flowchart TD
S1[Build acceptance checklist] --> S2[Audit six modules and landing]
S2 --> S3[Correct local drafts]
S3 --> S4[Mark open choices]
S4 --> D{Every comment resolved}
D -- yes --> DONE[Drafts aligned]
D -- no --> S2Design the smallest safe pilot operating model, including capacity, Rita workload, rhythm, video and Meet handling, consent, retention, success criteria, and readiness checks.
The pilot is described as small and manual, but its support load, data handling, readiness threshold, and timing are not explicit.
A premature date could create an unsustainable workload, unclear participant expectations, or unapproved recording practices.
Matt and Rita can select a date only after seeing the complete work, capacity, safety, and campaign preparation requirements.
A runbook contains capacity scenarios, responsibilities, service boundaries, consent and storage decisions, testing, and a backward schedule after approval. 80/20. One roughly ten-person operating scenario with explicit gates.
flowchart LR A[Small pilot idea] --> B[Bounded runbook] --> C[Known workload and risk] --> D[Date decision]
flowchart TD
S1[Model workload] --> S2[Define boundaries and data rules]
S2 --> S3[Add tests and success criteria]
S3 --> S4[Build backward timeline]
S4 --> D{All launch gates pass}
D -- yes --> DONE[Date can be chosen]
D -- no --> S2Prepare two interview invitation drafts, a Reservio-service specification, a Meet recording test, an interview guide, and separate optional marketing consent.
The call promised outreach, but recipient groups, scheduling, recording quality, reward eligibility, and public-use permission are not packaged.
Outreach remains delayed or risks treating a reward as conditional on public testimonial permission.
Rita receives a ready-to-approve workflow for honest feedback and optional testimonial material without conflating the two.
Two recipient-specific emails, a booking brief, Meet test, guide, and separate consent model are ready for review. 80/20. Two emails, one booking brief, one interview outline, and one two-step consent model.
flowchart LR A[Interview promise] --> B[Safe outreach workflow] --> C[Feedback and optional proof]
flowchart TD
S1[Draft two emails] --> S2[Specify booking service]
S2 --> S3[Test recording path]
S3 --> S4[Separate reward and consent]
S4 --> D{Workflow safe and clear}
D -- yes --> DONE[Package ready]
D -- no --> S1Turn the workshop idea into a bounded post-challenge private-beta brief for a small known audience, with a live format, learning objective, feedback, and continue or stop criteria.
The workshop has strategic potential but could distract from the challenge if treated as a parallel launch.
The idea is either forgotten or repeatedly reintroduced into challenge work without validation.
The next offer has a disciplined learning path while the challenge stays the priority.
A post-challenge brief defines invitee criteria, outcome, live format, boundaries, feedback, prerequisites, and decision criteria. 80/20. One private-beta hypothesis and feedback gate.
flowchart LR A[Future workshop idea] --> B[Private beta brief] --> C[Evidence to continue or stop]
flowchart TD
S1[Ground the concept] --> S2[Define audience and outcome]
S2 --> S3[Draft feedback and criteria]
S3 --> S4[Place after challenge]
S4 --> D{Beta brief bounded}
D -- yes --> DONE[Roadmap ready]
D -- no --> S2Package the approved drafts into one client review flow that tells Rita what to review now, what choices belong in the next meeting, and why the date remains open.
Rita has many scattered review items and the agency also owes work, so a generic request for feedback will not close the decision gaps.
The next meeting may repeat old discussion, miss unresolved comments, or choose timing before scope and workload are clear.
Rita gets a concise approval path, and the next call can end with named decisions, owners, and a defensible preparation window.
The package contains the deck, review checklist, Figma status, asset links, pilot decisions, testimonial materials, workshop roadmap, Hungarian email draft, and agenda. 80/20. One deck link, checklist, unsent email, and decision agenda.
flowchart LR A[Scattered review requests] --> B[One approval package] --> C[Focused client meeting] --> D[Named next steps]
flowchart TD
S1[Collect verified artifacts] --> S2[Build Rita review checklist]
S2 --> S3[Draft email and agenda]
S3 --> S4[Verify claims and links]
S4 --> D{Package complete}
D -- yes --> DONE[Ready for human send review]
D -- no --> S1