Daily Brief — 2026-06-10

> Generated: 2026-06-11 07:02 > Model: gemini-2.5-flash > Tokens: 71,100 in / 8,609 out > Cost: $0.0158 > Preset: small_team

---

The Day in Brief

The most critical signal today is a RED system health failure for the AIOS Screenpipe Daily Summary. This means your automated activity logs may be stale or missing, directly impacting the accuracy and completeness of the intelligence gathered by your AIOS. This foundational issue needs immediate attention to ensure your system is reliably tracking your daily work.

Despite the system health alert, you led a highly engaging "STARS - Weekly Training" session focused on Claude. You provided a comprehensive overview of Claude's capabilities, models, and project features, emphasizing its role as a business partner rather than just a chatbot. A key highlight was the introduction of the "My Voice DNA" skill, designed to personalize Claude's output to match your unique writing style.

A notable challenge during the live training was the failure of the prompt generator artifact to load, causing a momentary setback. However, you adeptly navigated this, demonstrating the tool's functionality through a pre-recorded video and promising a fix or an alternative GPT. This incident underscores the "fun with AI when things don't work" and the need for robust product delivery.

A significant personal win was shared: you successfully built a CLI connector for Agent Opus, automating the creation of 45 videos in just five minutes, despite the tool lacking a public API. This impressive feat showcases your "Activator" strength and the power of Claude Code for complex automations. This month's focus on Claude training directly supports your strategic priority of growing AI education recurring revenue and reducing reliance on doTERRA income.

Today, watch for the resolution of the Screenpipe failure and prioritize the fix for the prompt generator artifact to maintain momentum with your Stars members.

Key Signals

🔥 Agent Opus CLI Automation: Sondra successfully built a CLI connector for Agent Opus, automating 45 video creations in 5 minutes, a significant personal automation win. ⚠️ AIOS Screenpipe Failure: The scheduled_tasks: AIOS Screenpipe Daily Summary is RED, indicating potential missing or stale activity data for your AIOS. ⚠️ Prompt Generator Artifact Failure: The prompt generator artifact failed to load during the live Stars training, impacting product demonstration. 💡 New Claude Models: Claude Fable 5 and Mythos 5 (with Ultra Code/dynamic workflows) are new top-ranked models, offering fresh teaching opportunities for Stars. 📊 Stripe MRR Growth: Monthly Recurring Revenue increased by $14.25 to $110.00, with 3 new active subscriptions. 📊 One-Time Revenue Spike: A significant $973.00 in one-time revenue in the last 7 days, primarily from 5 sales of the Monthly Drop - Video Vault. 📌 Stars Training Engagement: High member participation and excitement during the Claude training, with members sharing their personal AI wins. 💡 Multilingual Voice DNA Opportunity: Zully's question about creating separate Voice DNAs for formal Spanish and English highlights a potential niche for specialized training. 📌 Activator Drift Risk: Sondra's excitement about the Agent Opus CLI automation, while impressive, could lead to new projects that divert focus from the current priority stack. 🔥 Voice DNA Skill Creation: Successful demonstration and creation of the "My Voice DNA" skill, empowering members to personalize Claude's output.

Metrics Analysis

System Health: The scheduled_tasks: AIOS Screenpipe Daily Summary is flagged RED with an exit code of 267014. This is a critical issue as it suggests that the automated daily summary of your Screenpipe activity is failing. This means the AIOS system may not have up-to-date or complete data on your recent work, which directly impacts the accuracy of this brief and your overall intelligence layer. This requires immediate investigation.

YouTube Channel Performance: Your YouTube channel currently has 1,060 subscribers, 62,253 total views, and 163 total videos. In the last 30 days, only 1 video was published, generating 18 views. This indicates a very low recent content cadence and engagement, which is below average compared to a channel actively trying to grow. This aligns with your recent decision to revamp the channel via Sandra Lee's program and intentionally keep the posting cadence loose, acknowledging that it had fallen by the wayside. The current numbers reflect this strategic pause and rebuilding phase.

Stripe Revenue:

• Monthly Recurring Revenue (MRR): Your MRR stands at $110.00, representing a healthy increase of $14.25 from last week's $95.75. This positive trend is supported by 3 new active subscriptions, bringing your total to 22, up from 19 last week. This indicates steady, albeit modest, growth in your recurring revenue streams.
• Revenue Month-to-Date (MTD): You have generated $1000.00 in revenue MTD.
• New Subscriptions: 4 new subscriptions were acquired this month.
• Cancellations and Churn: There was 1 cancellation, resulting in a churn rate of 4.5%. This is a manageable churn rate, especially with new subscriptions offsetting it.
• One-Time Revenue (Last 7 Days): This is a significant highlight, with $973.00 generated from 13 transactions, a substantial increase from last week's $44.00. The primary driver was 5 sales of the Monthly Drop - Video Vault, each at $147.00, totaling $735.00. This strong performance in one-time sales directly validates the value of your AmplifAI Video Studio and the Agent Opus challenge, demonstrating clear market demand for these video assets.

Skool Affiliate Earnings: Lifetime affiliate earnings are $2,282.89, an increase of $22.50 from last week's $2,260.39. This shows consistent, albeit smaller, passive income generation from your 36 active referrals. While not a primary growth engine, it's a stable supplementary revenue stream.

Overall Analysis: The financial metrics show strong positive momentum, particularly in one-time sales driven by the AmplifAI Video Studio. This validates your efforts in the Agent Opus challenge and provides a clear path for further monetization of video assets. The growth in MRR and active subscriptions is encouraging for your core AI education business. The YouTube performance is low but aligns with a deliberate strategic shift. The critical concern remains the AIOS Screenpipe failure, which could mask other operational issues if not resolved promptly.

Meeting Highlights

CALL 1: STARS - Weekly Training | 2026-06-10

• Key Decisions Made:
• The current month's Stars training will be entirely focused on Claude, starting with fundamental capabilities like models, memory, projects, and voice DNA.
• Sondra will provide members with a prompt generator artifact and a "My Voice DNA" skill.
• Homework for members includes importing their memory into Claude and building their personal Voice DNA.
• Future training will cover Claude Co-work (later), a skill tester (coming soon), and a content pipeline.
• Sondra will create a GPT for the prompt generator if the artifact cannot be fixed.

• Action Items:
• [Sondra] Provide access to the Claude overview PDF in the Stars area.
• [Sondra] Provide the prompt generator artifact link (after fixing the loading issue).
• [Sondra] Provide the "My Voice DNA" skill file to Stars members.
• [Sondra] Investigate and fix the prompt generator artifact loading issue that occurred during the live session.
• [Sondra] Share Nate Herc's CLI video with "experts" in the community as promised.
• [Sondra] Follow up with Zully on her question regarding creating separate Voice DNAs for formal Spanish and English.

• Notable Signals:
• Sondra's Automation Win: Sondra shared a significant personal achievement: building a CLI connector for Agent Opus (a tool without a public API) and automating the creation of 45 doTERRA product videos in approximately five minutes. This demonstrates advanced AIOS application and is a powerful testament to the "Build with AI" core identity.
• High Member Engagement: Members like Tina ("I built my website yesterday with Claude") and Kim ("Claude is my new best friend") expressed high enthusiasm and shared their own successes with Claude, indicating strong resonance with the training.
• Personalized AI: Sondra noted her Claude quoted scripture back to her, illustrating the deep level of personalization and memory Claude can achieve over time.
• Multilingual Opportunity: Zully's question about handling formal Spanish and English voices in Claude highlights a potential niche for future content or skill development.
• Technical Glitch: The prompt generator artifact failed to load during the live demo, causing a minor disruption but was handled by showing a pre-recorded video.

• Sara Dependencies: None. As per the updated business context, Sara is on hold, and no commitments or open items related to her were identified in this meeting.

Strategic Recommendations

1. Concerns:

1. AIOS Screenpipe Failure: The RED status of scheduled_tasks: AIOS Screenpipe Daily Summary is a critical operational concern. This failure means your AIOS is not reliably capturing and summarizing your daily activity, directly compromising the "Data" and "Intelligence" layers of your system. Without accurate activity logs, the insights provided by this brief and your overall AIOS become less trustworthy. Recommendation: Immediately investigate the Last exit code 267014 and resolve the Screenpipe summary task to restore full data integrity.
2. Prompt Generator Artifact Reliability: The live failure of the prompt generator artifact during the Stars training is a significant product delivery issue. It undermines the "Implementation" layer of your ecosystem and can erode member trust. While you handled it well, consistent product functionality is paramount. Recommendation: Prioritize diagnosing and fixing the artifact's loading issue. If a quick fix isn't feasible, immediately implement your contingency plan to create a GPT alternative, ensuring members have access to this promised tool without further delay.
3. Activator Drift Risk from Agent Opus CLI: Your impressive achievement of automating Agent Opus videos via a CLI connector is a testament to your "Activator" strength. However, the excitement around this new capability could lead to a new, unplanned teaching or productization initiative that diverts focus from your established Active Priority Stack (AI Asset Weekend Builder, AI Marketing Team). This risks "parallel build overload" and delays on revenue-generating priorities. Recommendation: Acknowledge this win, but consciously categorize it as a personal automation success for now. Resist the urge to immediately productize or teach it until the AI Asset Weekend Builder and AI Marketing Team launches are stable and generating revenue.

2. Opportunities:

1. Capitalize on AmplifAI Video Studio Momentum: The significant spike in one-time revenue ($973.00 in 7 days), largely driven by the Monthly Drop - Video Vault sales, clearly indicates strong market demand for AI-powered video assets. This validates your efforts in the Agent Opus Challenge. Recommendation: Accelerate the launch of your DFY Video Service (Priority 4) and intensify follow-ups for the "Biz in a Box" international founders, leveraging this proven demand. Ensure the Montage Agent and brand-consistent video features from Agent Opus are integrated into your Stars demos and marketing.
2. Leverage New Claude Models for Stars Content: The recent release of Claude Fable 5 and Mythos 5, particularly with the "Ultra Code" dynamic workflows, presents a timely and high-value opportunity for your Claude Lab Training inside Stars (Priority 5). This cutting-edge information can significantly enhance your curriculum and attract new members. Recommendation: Integrate Fable 5 and Ultra Code into your upcoming Claude Lab sessions, demonstrating their capabilities and practical applications. Verify pricing and official details from Anthropic before teaching.
3. Address Multilingual AI Voice DNA: Zully's question about creating separate Voice DNAs for formal Spanish and English highlights an underserved market segment. As an AI Educator, providing solutions for multilingual entrepreneurs could differentiate your offering. Recommendation: Explore developing a specialized skill or a dedicated training module within Stars for creating and managing multilingual AI voices, potentially expanding your reach to a broader international audience.

3. Quick SWOT:

• Strength: Sondra's exceptional ability to build and automate complex AI solutions (e.g., Agent Opus CLI) and her highly personalized, engaging approach to community education.
• Weakness: Critical system health failures (Screenpipe) and product delivery issues (prompt generator artifact) are undermining the reliability and perceived value of the AIOS and offerings.
• Opportunity: Strong market validation for AI-powered video content and the timely release of advanced Claude models create immediate avenues for revenue growth and cutting-edge education.
• Threat: The inherent "Activator drift" pattern, combined with current overwhelm, risks diverting focus from completing existing, revenue-generating priorities in favor of new, exciting builds.

Action Items

1. [Sondra] Investigate AIOS Screenpipe Failure: Diagnose and resolve the scheduled_tasks: AIOS Screenpipe Daily Summary RED failure (Last exit code 267014). This is critical for restoring data integrity to your AIOS. (ASAP)
2. [Sondra] Fix Prompt Generator Artifact: Prioritize fixing the prompt generator artifact that failed to load during the Stars training. If a quick fix is not possible, create the planned GPT alternative immediately. (ASAP, before next Stars session)
3. [Sondra] Distribute Claude Overview PDF: Make the Claude overview PDF accessible to Stars members in the designated area. (Today)
4. [Sondra] Distribute "My Voice DNA" Skill: Ensure the "My Voice DNA" skill file is readily available for Stars members to download and implement. (Today)
5. [Sondra] Share Agent Opus CLI Video: Share Nate Herc's CLI video with the "experts" in the community as promised during the training. (Today)
6. [Sondra] Follow up on "Biz in a Box" Founders: Follow up with all international founder prospects (Jackie, Denise, Manu, Azaleea Zaman) before their Sunday (June 8) decision deadline to convert sales. (ASAP, before June 8)
7. [Sondra] Plan Stars Demo of Montage + Branding: Schedule a Stars demo of Agent Opus's Montage Agent and brand-consistent video features the week they drop, showcasing their practical application. (This week)
8. [Sondra] Test Multi-Brand Setup for doTERRA: Test the multi-brand setup for doTERRA within Agent Opus to ensure it supports your specific use case for the video library. (This week)
9. [Sondra] Address Multilingual Voice DNA: Explore Zully's question about creating separate Voice DNAs for formal Spanish and English, potentially as a new skill or training module. (Next Claude Code session)
10. [Sondra] Review Gemini Article Quality: Read the first draft article in the Digital Home dashboard ("How to Use AI to Write a Weekly Newsletter in 15 Minutes") and provide feedback on Gemini's output quality against your brand voice. (This week)

---

Your Open Questions

Here are the answers to your open questions, drawing from the compiled business intelligence:

Claude Skills Repo

• Track ~/.agents/skills/ too? It is recommended to fold the unique skills from ~/.agents/skills/ into your existing digitalbeachgirl/claude-skills private GitHub repo. This centralizes all your custom skill IP under version control and ensures consistent backup, rather than relying solely on Backblaze for a partial mirror.

Agent Opus CLI

• Photo upload not built: This requires a dedicated Claude Code session to build the upload flow. The necessary endpoints are documented in your (gitignored) builds/agent-opus/ specs.
• Monthly Drop scripts source: A decision is needed on whether to point the video helper at the FINAL doTERRA Excel file or a folder of .txt files for batching videos.
• Read-command rough edges: This is a minor issue. A Claude Code emboss pass on the agent-opus-pp-cli can generalize the get-by-id commands that currently have a captured project ID baked in.

Monthly Drop Club V3 scripts

• Blog enrichment pending for 90 of 108 scripts: RESOLVED (2026-06-08). The full blog-enrichment pass over all 90 product-page-only singles and blends was completed overnight. All 108 scripts now cite a doTERRA-owned blog in V3 Sources.
• MetaPWR product URL unconfirmed: RESOLVED (2026-06-08). The correct product page and blog URLs for MetaPWR were confirmed and written to row 90 of your Excel.

Digital Home - Safety Check Findings

• RED #1 - No privacy policy (frontend): This is a legal must-fix. The site collects email and drops tracking cookies without a /privacy page, posing Stripe/ad-platform ban risks.
• RED #2 - No cookie consent banner (frontend): This is also a legal must-fix. Four tracking cookies are set automatically before interaction, requiring consent for EU/CA compliance.
• YELLOW #3 - Article HTML not sanitized (frontend, XSS): This needs a Claude Code session to implement DOMPurify on article.body to prevent Cross-Site Scripting vulnerabilities.
• YELLOW #4 - Admin board goes blank on failed load (backend): This needs a Claude Code session to add an app/error.tsx boundary to handle error states on content page fetches.
• YELLOW #5 - Cloudflare env vars: You need to confirm that NEXT_PUBLIC_SITE_NAME, NEXT_PUBLIC_SITE_URL, and RESEND_FROM_ADDRESS are correctly set in your Cloudflare environment variables to prevent branding leaks and email bounce issues.
• Lawyer later: This is a note for future consideration if you scale ads into Europe, to have a privacy lawyer review.

Monthly Drop Club Script Pipeline

• Lip Balm (#164): RESOLVED. A single collection script covering all three flavors was created.
• Script length: RESOLVED. All product scripts are now targeted for 1 minute to 90 seconds (~130-200 words).
• Encoding corruption: RESOLVED. The Excel file was cleaned to pure ASCII, removing special characters.
• Scripts need full rerun (URL only): You need to confirm which scripts you manually rewrote, then rerun the remaining scripts using only the product URL as the source, not old script drafts.
• Excel file locked: The ~$FINAL doTERRA_All_Products_2026_with_scripts.xlsx lock file indicates the Excel file is currently open. You must close and reopen the file before any changes can be saved.

doTERRA Video Factory

• Viral short-form script skill not yet uploaded: RESOLVED (2026-06-01). The viral-short-form-script-engine skill was built and installed globally, with a shareable version packaged.

Ikigai — 3 unanswered questions These are deep personal reflection questions for you, Sondra, and Claude cannot answer them directly. They require your personal introspection and strategic decision-making.

• AI or just create-and-teach?
• YouTube gap: foundation or publishing fear?
• Shared-cost business: real desire or sunk-cost rescue?

Financial / Billing

• Anthropic API billing change: This needs verification. Check your Anthropic billing dashboard and account settings to confirm if programmatic (API) usage no longer counts toward your $200/mo Claude Max plan, as this could significantly impact AIOS automation costs.

Infrastructure / Backup

• Sondra has NO laptop backup outside git: RESOLVED (2026-06-01). You confirmed that Backblaze performs a full, whole-computer backup, including all gitignored content. No further action is needed.
• Token expiration unknown: RESOLVED (2026-06-01). Your GitHub PAT has no expiration, as confirmed by the GitHub API and your recollection. No calendar reminder is needed.
• Misconfigured remote in PROJECTS/Pushten website/amplifai-template-studio/: This needs a quick check via Claude Code to confirm if the origin remote pointing to AmplifAI-Your-Business/test-sondra-blog-website.git is an intentional clone or a stale misconfiguration.

Base44 Blog Template — Productization Open Items

• Apps 2 and 3 not yet replicated: This requires a Claude Code session to apply the usePageSeo hook, all dashboard toggles, and meta description fixes to these apps.
• Prerender.io vs Vercel migration vs marketing reposition decision pending: This is a strategic decision for you, weighing the costs and benefits of each path for full AI-search visibility.
• og:image default still Base44 logo: This needs a Claude Code session to identify where the platform-wide default OG image is set in the pre-React HTML and update it to your branding.
• 0 Content Types and 0 Niches defined on master: This needs a Claude Code session to define content types and niches, and to inspect the pre-built templates to confirm they produce AEO/GEO content.
• Owner Playbook v2 acceptance: You need to review and explicitly accept or reject the v2 PDF of the Owner Playbook.
• Customer-facing Owner Playbook needs updating: This requires a Claude Code session to update the Owner Playbook to reflect the verified state, buyer-required steps, and honest Base44 limitations.
• Publish UX wall: A decision is needed: either document the STATUS dropdown → Published process loudly for non-technical buyers, OR send a Base44 editor prompt to replace it with clear [Save Draft] [Publish] buttons.
• AI Write Post voice quality: This needs either a voice-training intervention for the AI or clear positioning to buyers that the feature produces "starting drafts only" to manage expectations and prevent refunds.

Blog Owner Playbook

• Owner Playbook navigation rebuild: RESOLVED (2026-05-23). V2 was shipped with a clickable TOC, lookup table, hyperlinks, and Site Config crops.
• Visual review of v2 PDF: You need to open the v2 PDF, click 3-5 random Lookup Table rows to confirm hyperlinks, and skim the Site Config crops to verify accuracy. If crops are off, adjust scripts/crop_site_config.py and rebuild.
• Length trim after navigation lands: This decision is deferred until you have used the v2 PDF, as improved navigation may make the page count feel less overwhelming.
• Document B (Client Guide): This remains deferred.
• Site Config crops — auto-crop vs hand-crop: A decision is needed if the auto-crop bands in v2 appear cut off; you can hand-crop with Snipping Tool and rebuild if necessary.

SEO Blog Website Template — Productization

• Modal readability fix verification: You need to visually verify the .admin-white-card className fix in AdminModal.jsx after the most recent deploy.
• pSEO + indexing comprehensive fix deployment status: This requires end-to-end verification of the multi-feature prompt (auto-images, OG Images, Silo Links, sitemap, IndexNow, Featured Image Picker, dashboard cards) across all three apps.
• Master Base44 app selection: This is a strategic decision for you to select which of the three Base44 apps will serve as the production master for cloning to new buyers.
• Pricing tier formal confirmation: You need to formally accept or adjust the proposed pricing tiers ($1,500 + $97/mo, $3,500 + $197/mo, $7,500 + $497/mo).
• First buyer / launch path: This is a strategic decision for you to pick the first client target (e.g., doTERRA international partner, GEO Agency lead, Stars member).

Digital Home Website Service — Scope to Lock

• Pricing confirmation: You need to accept or adjust the proposed pricing for the three tiers.
• Monthly management scope per tier: You need to accept or adjust the proposed scope of monthly management for each tier.
• Stripe model for e-commerce clients: A decision is needed: manual payout for v1 or set up Stripe Connect from day one? (Default is manual until a client needs e-commerce).
• Sara's role: Given Sara is on hold, the default is for you to handle client #1-3 solo.
• First client target: This is a strategic decision for you to pick the first client (e.g., doTERRA international partner, GEO Agency med-spa lead, Stars member).

Infrastructure

• Wrangler CLI not authenticated locally: This requires running wrangler login (browser flow) when convenient to enable wrangler deploy and wrangler secret put locally.

Digital Home Backend — Gemini swap + UI gaps

• Gemini output quality vs Claude: You need to read the first draft article in the dashboard ARTICLES tab ("How to Use AI to Write a Weekly Newsletter in 15 Minutes") and judge whether Gemini's writing matches your brand voice.
• Where was the original unauthorized Claude swap Sondra remembered?: This needs investigation across Pushten website, Amplif-AI Template Studio, Coach Blog Template, and BraveBrand intelligence app.
• Backend dashboard UI gaps: This needs a dedicated UI improvement session via Claude Code to address issues like missing buttons, placeholder modules, dark theme readability, and sidebar scroll bugs.
• No POST handler on /api/content-calendar: This needs a Claude Code session to add a POST handler to enable the future "Add Topic" button.
• package-lock.json has 122 unrelated deletions: This needs verification or discarding before the next commit to the backend repo.
• Trend scan uses Google News RSS — blocked by Cloudflare IPs: RESOLVED (2026-05-19). The /api/trend-scan was swapped to Bing News RSS.

Template-Blog-website Migration

• Phase 3 shim unverified: RESOLVED (2026-05-19). All 13 Supabase tables were built, RLS applied, and the shim patched and deployed.
• Guessed column types in 0001_initial_schema.sql: Monitor for errors on specific admin pages (Pillars, Categories, etc.); column types may need adjustment if misbehavior occurs.
• Phase 4 or 5 next?: This is a strategic decision for you to choose between implementing backend functions (Phase 4) or AI content via Claude API (Phase 5) next.
• Phase 4 — Backend functions still needed: This requires a Claude Code session to build publishScheduledPosts, generateSitemap, and weeklyReport using Supabase Edge Functions + Resend.
• Phase 5 — AI content generation pending: This needs a Claude Code session to integrate the Anthropic Claude API into AdminGenerate.jsx and AdminNewPost.jsx.
• Phase 6 — Media uploads + Supabase Storage: This needs a Claude Code session to configure media uploads and Supabase Storage.
• GitHub auto-deploy to Vercel not connected: This requires granting org-level permission to the Vercel GitHub app for AmplifAI-Your-Business in Vercel settings.
• Custom domain for template-blog-website.vercel.app: A decision is needed on whether to configure a custom domain for this site.
• Resend account for Phase 4 emails: A decision is needed: reuse Digital Home's Resend account or create a separate one for Phase 4 emails.
• TweetGod Supabase restore timing: A decision is needed on when to restore the TweetGod bot's Supabase project.

Amplif-AI Template Studio

• Purchase links: what payment mechanism?: This is a strategic decision for you to choose between Stripe direct links, GHL checkout, or ThriveCart for your "Buy Now" buttons.
• Custom domain for template studio?: A decision is needed on whether amplifai-template-studio.vercel.app should get templates.amplifaiyourbusiness.com or its own domain.
• GitHub connection to Vercel not linked: This requires granting access to the AmplifAI-Your-Business org in Vercel settings to enable auto-deploy on push.
• TweetGod Supabase paused: A decision is needed on when to restore the TweetGod bot's Supabase project.

Digital Home Backend

• Railway vs Cloudflare Workers: RESOLVED (2026-05-20). Cloudflare Workers are being used.
• OpenAI API key: RESOLVED (2026-05-20). The key is confirmed and set as a Worker secret.
• Trend scan uses Google News RSS — blocked by Cloudflare IPs: RESOLVED (2026-05-19). Swapped to Bing News RSS.
• Backend GitHub repo is private: RESOLVED (2026-05-20). The repo is private, and Cloudflare CI is connected.
• Frontend pages not yet branded: This needs a BraveBrand Prompt Builder session to generate a redesign prompt, then a Claude Code session to apply it to the Digital Home frontend project.

Brand Identity

• context/brand-detail.md not yet updated: This needs a Claude Code session to load brand-detail and apply the updates from your May 2026 ICA and Brand Identity documents.

GEO + Reputation Agency

• Zubair's AI GEO Guide v3: GOT IT (2026-05-18). All new intel has been added to PROJECTS/GEO-Agency/ docs.
• Service domain strategy: This is a strategic decision for you to determine if the GEO/reputation service will live on sondraverva.com or a separate brand domain.
• Self-serve app brand name: A decision is needed on the brand name for the self-serve app (e.g., "GEO Score," "AI Visibility Score," or Sondra-branded).
• Jonathan Mast's GEO app: This needs research to understand its functionality and whether it's competitive or complementary to Zubair's tool.
• Sara + Sondra legal agreement: This is on hold as Sara is on hold.

Digital Home (SondraVerva.com)

• Content corpus needs customization: RESOLVED (2026-05-18). All 10 files were populated from your Obsidian brain, BraveBrand Intelligence App, and Skool testimonial vault.
• brand_context Supabase rows not yet added: This needs a Claude Code session to add the three required rows to the brand_context Supabase table.
• Pages not yet branded: This needs a BraveBrand Prompt Builder session to generate a redesign prompt, then a Claude Code session to apply it to the Digital Home project.
• GitHub Actions need SITE_URL variable: This requires the backend to be deployed and the SITE_URL repo variable to be set in GitHub Settings.
• Backend setup pending: This needs a separate Claude Code session with the PROJECTS/DIGITAL HOME V1/Digital_home_backend_starter/ folder as root.
• Resend domain verification pending: Before going live, verify sondraverva.com in the Resend dashboard and update RESEND_FROM_ADDRESS in .env.local.
• RESEND_WEBHOOK_SECRET is placeholder: This is not blocking, as it's only needed for email webhook events.
• Next.js workspace root warning: This is not urgent, but can be fixed by adding turbopack.root to next.config.ts.

Claude Code Setup

• Agent teams not yet enabled: This requires manually adding "env": { "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1" } to ~/.claude/settings.json. You should verify Anthropic billing first, as agent teams can cost 4x solo sessions.
• Delete duplicate skill files from module-installs/: You can safely delete module-installs/youTube_search_skill.md and module-installs/youtube_pipeline_skill.md as they are duplicates of already-installed skills.

Skool Admin Dashboard

• Dashboard posting not yet confirmed end-to-end: This needs diagnosis. Paste the full terminal output from the Flask server to identify the issue and ensure SKOOL_EMAIL and SKOOL_PASSWORD are correctly set in .env.
• Post as Sondra or as AI Admin?: A decision is needed on whether to post as your main Sondra account or the "Sondra's AI Admin" account.

AIOS Dashboard — Obsidian Integration

• Custom Frames vs full custom Obsidian plugin?: This is a strategic decision for you, weighing the 30-minute Custom Frames plugin (works today) against a polished, custom Obsidian plugin (1-2 sessions).
• INSTALL-AUTO-LAUNCH.bat not yet run: This needs to be run as administrator to enable the dashboard to auto-launch at login.
• Notion Task Scheduler triggers not yet disabled: RESOLVED (2026-06-07). You decided to keep Notion alive as a live downstream mirror, so these tasks remain active.
• ffmpeg + Groq key ready (2026-05-17): After a Claude Code restart, run /watch on any YouTube URL; the skill's setup wizard will confirm or prompt for the Groq key location.

⚠️ Anthropic Pricing Change

• Programmatic (API) usage no longer counts toward $200/mo Claude Max plan: This needs verification. Check your Anthropic billing dashboard and account settings to understand the impact on AIOS automation costs.

Thumbnail Generator

• Photo catalog not yet set up: FULLY RESOLVED (2026-05-15). 40 photos were classified, 4 scripts built, and the generator confirmed working.
• 10 Facebook photos unconfirmed: This needs visual confirmation and proper expression/pose classification for the 10 Facebook photos in your catalog.
• search_examples.py yt-dlp path bug: This needs a Claude Code session to fix the yt-dlp path bug by using shutil.which('yt-dlp').
• Gemini face likeness quality: Evaluate the likeness quality after a few more generations. If insufficient, the fallback is to generate without a face and composite your real photo in Canva.
• Scrape Creators API credits: Monitor your usage on the $29/mo plan.

YouTube Project

• Full YouTube workflow not yet built: This needs a Claude Code session to build the complete system, starting with competitor research.
• Competitor research pending: This needs a Claude Code session to find 3 AI educator channels at ~1k subscribers, pull top thumbnails, and audit them against Lyndon's 3-element framework.

Base44 Community Site

• Which Amplif-AI portal app to develop further?: This is a strategic decision for you to pick between 69fdb3b330941425ae20e00b (navy blue theme) and 6a020afa3d93fd72236a8dad (VCS-style clone).
• Lovable MCP not yet tested: It is worth testing https://mcp.lovable.dev on a landing page build.

AIOS / Skool Tools

• Daily cross-post community URLs: You need to provide the slugs for the Skool communities you post to daily to complete the cross-posting setup.
• Weekend hourly posting queue: This needs content before the scheduler can be built.
• Higgsfield CLI: You need to run higgsfield auth login to activate it and decide on its intended use.
• Skool category IDs: The other category IDs need to be mapped beyond the General feed.

AIOS / Data Integrity

• What is the actual current total Skool member count?: This needs verification and an update to context/current-data.md.
• Where did "Skool rank #1824" come from?: This needs confirmation to clarify if it refers to overall Skool rank versus Tech category rank.
• May 8–9 Screenpipe data permanently lost: This is noted for reference; no recovery is possible.

Tools & Privacy

• Littlebird.ai — worth trying?: Evaluate the free tier for one week to see if it complements Screenpipe for meeting notes.
• OpenAI Privacy Filter model — upgrade later?: This is deferred; revisit if the current regex misses important PII in practice.

Tools to Evaluate

• Claude Memory Compiler: This is worth evaluating as it could enhance or replace your manual Obsidian brain compile.
• Context Engineering Intro: Compare this planning workflow (INITIAL.md → /generate-prp → /execute-prp) with your existing /plan-feature before deciding on installation.
• Open CoDesign first run: This needs API key setup on first launch (select Local CLI mode) and practical testing.

Income / Financial

• Other affiliate programs (non-Skool): It's worth checking bank statements for unidentified income sources from other affiliate programs.

AIOS / Claude Code (general)

• Will Claude Code see retroactive edits to daily notes?: Claude Code reads the compiled brain. If notes are compiled, it will see them regardless of when they were originally written.
• **Is Screenpipe data actually visible to