Projects · 16

1
Shipping
5
Building
2
Validating
3
Concept
5
Shelved
6 / 2
Active commits (D-002 cap)

Tasks · 106 open

14
Blockers (cross + per-project)
37
High-severity
7
Cross-portfolio (registry)
99
Per-project (FOUNDER-TASKS.md)
106
Total open

Top 5 cross-portfolio (registry)

  • Blocker Compliance Hintly compliance gate — DPAs + EOOD + sub-processor disclosure T-005 due in 5d
  • High Financial Reconcile tax tracking — pick one source of truth T-002 due in 21d
  • High Strategic Decide which 2 of the 6+ concept projects to keep alive T-001 due in 51d
  • Medium Business ClinicFlow validation — outreach to 20 BG clinics T-006 due in 21d
  • Medium Business List at least one Spreadsheet OS on Etsy T-004 due in 36d

Top 5 per-project (FOUNDER-TASKS.md)

  • Blocker Set `EOOD_REGISTERED_ADDRESS` in Vercel env — every transactional email currently shows a placeholder. CAN-SPAM Act §316 LB1 · hintly
  • Blocker EOOD formation finalized LB9 · hintly
  • Blocker Liability insurance LB10 · hintly
  • Blocker Counsel retention LB11 · hintly
  • Blocker Verify Vercel env vars all set — full list below in 🟠 section. `ANTHROPIC_API_KEY`, `SUPABASE_*`, `STRIPE_*`, `RESEND_AP LB12 · hintly

Upcoming · next 30 days

  • in 5d 2026-05-15 Compliance gate (DPAs + EOOD + sub-processor disclosure) — Hintly
  • in 12d 2026-05-22 Soft launch — named list of 30 — Hintly
  • in 21d 2026-05-31 20 BG clinics outreach complete — AI Agent (ClinicFlow)
  • in 22d 2026-06-01 Public launch — Hintly

Top risks · active projects

  • Hintly Shipping
    COPPA + sub-processor disclosure consistency before first paid customer.
  • AI Agent (ClinicFlow) Validating
    Building before validating willingness to pay. The PRD assumes the customer exists.
  • Tax Project Building
    Three places tracking the same numbers (project + 2 spreadsheets).
  • Spreadsheet OS collection Building
    9 SKUs with no listing is 0 listings. Pick a winner.
  • MDM Engine (Petko Digital Products Ideas) Building
    Overlap with Pet+Owner OS unclear.
  • Freelancer OS / SoloistOS Building
    Versioning forever without launching. v4 → v5 → v6 with no v1 in market.

Recent activity

Recently updated projects

  • DreamQuest — Bedtime Adventures Shelved · 0d ago · next:
  • Pet+Owner Operating System Concept · 0d ago · next: Decision needed — fold into T-001 (which 2 of 7+ concepts to keep alive). Currently no active invest
  • PetkoOS — Execution Engine ("Finish What Matters") Building · 0d ago · next: No active work until trigger fires. Re-evaluate at 2026-06-01 (Hintly launch) and 2026-08-01 (Hintly
  • AI Help Assistant (productized service) Concept · 0d ago · next: No active work until trigger fires. Material preserved.
  • Hintly Shipping · 1d ago · next: Execute the launch checklist (FOUNDER-TASKS.md). Compliance gate before paid promo.
  • LinguaLeap Shelved · 1d ago · next:

Recent decisions

  • D-003 Mobile access via claude.ai Project bundle (Path B); Cowork (Path C) deferred · 2026-05-10
  • D-004 ProjectA (AI Help Assistant) deferred — product-only portfolio strategy · 2026-05-10
  • D-005 PetkoOS (Execution OS) deferred — Hintly post-launch trigger · 2026-05-10
  • D-006 DreamQuest shelved — Hintly success-tied revival trigger · 2026-05-10
  • D-001 Personal OS becomes the canonical home for cross-project state · 2026-05-09

By project

One panel per project — status, KPIs, milestones, tasks, FOUNDER-TASKS, spinoff scan, P&L scenarios, docs. Filter by status AND/OR jump to a specific project.

Status:
Project:

Hintly

Visual-first hint product for kids learning. K-8.

kidL&CAICOPPAK-8B2C
Shipping P1 · this week
Stage
pre-launch sprint
Market
US (kid learning, parent-purchased)
Model
Free + paid subscription (€/$ TBD)
Path
Hintly
Last update
2026-05-09
Next action
Execute the launch checklist (FOUNDER-TASKS.md). Compliance gate before paid promo.
Biggest risk
COPPA + sub-processor disclosure consistency before first paid customer.

Open tasks (cross-portfolio: 1)

  • Blocker Compliance Hintly compliance gate — DPAs + EOOD + sub-processor disclosure T-005 due 5d doing

FOUNDER-TASKS.md (56 open)

Blocker Blocker (6)
  • LB1 Set `EOOD_REGISTERED_ADDRESS` in Vercel env — every transactional email currently shows a placeholder. CAN-SPAM Act §316.4(a)(5) requires valid postal address. PR #66 audit pass 2 / `_emailLayout.js`
  • LB9 EOOD formation finalized Memory `project_launch_plan_2026_05.md` B-list
  • LB10 Liability insurance Memory B-list
  • LB11 Counsel retention Memory B-list
  • LB12 Verify Vercel env vars all set — full list below in 🟠 section. `ANTHROPIC_API_KEY`, `SUPABASE_*`, `STRIPE_*`, `RESEND_API_KEY`, `EOOD_REGISTERED_ADDRESS`, etc. PR #66 + .env.example
  • LB13 Common Sense Privacy submission Memory B-list / `docs/compliance/common-sense-privacy-submis
High High (20)
  • EM1 Confirm `support@hintlyapp.com` routes to founder inbox always
  • EM2 Confirm `team@hintlyapp.com` routes (referenced in /careers) PR #65 audit S6
  • EM3 Confirm `privacy@hintlyapp.com` routes (referenced in privacy + KB) PR #65 audit S6
  • EM4 SPF / DKIM / DMARC for hintlyapp.com → Resend (deliverability) always
  • EM5 Test welcome email lands in Gmail / Outlook / iCloud inbox (not spam) implicit
  • SO1 Claim `@hintlyapp` on X (or remove footer icon) PR #65 audit S7
  • SO2 Claim `@hintlyapp` on Instagram (or remove icon) PR #65 audit S7
  • SO3 Claim `@hintlyapp` on LinkedIn company page (or remove icon) PR #65 audit S7
  • OP1 First eval-suite run against launch baseline → commit `eval/quality-suite/results/baseline-2026-05.md` PR #59 STRESS-TEST-PLAN + PR #63 LEARNING-LOOP
  • OP2 First k6 stress-test run against a Vercel preview → commit `docs/STRESS-TEST-RESULTS.md` PR #59
  • OP3 Lighthouse-CI baseline numbers committed (after Lighthouse runs in PR #47) PR #47
  • OP4 Healthchecks.io account → set `HEALTHCHECKS_PING_URL` env var Memory B-list / Agent playbook 5.10
  • OP5 Slack workspace incoming webhook → set `SLACK_ALERT_WEBHOOK_URL` env var Agent playbook Phase 1
  • OP6 PostHog read-only API key → set `POSTHOG_API_KEY` for daily founder digest agent Agent playbook 9.2
  • OP7 PostHog feature-flag config (`signup-trial-variant`, etc.) reviewed and shipped Memory B-list
  • OP8 Decide founder-digest email destination + send time (default 7am EST to founder) Agent playbook 9.2
  • OP9 GSC (Google Search Console) validate-fix on any pre-launch crawl errors Memory B-list
  • OP10 Verify Supabase `pgcron` jobs scheduled correctly (not skipping) implicit
  • OP11 Test the full signup → onboarding → first session → weekly summary flow end-to-end with a clean test account implicit
  • OP12 Send a test welcome email + weekly digest to yourself; confirm no placeholder text PR #66 / this PR
Medium Medium (6)
  • M1.1 Real testimonials collection → replace placeholder spots with quotes Memory `project_overhaul_plan.md`
  • M1.2 Wire 8 Phase-1 agents from agent playbook (incident response, daily digest, error budget, COPPA audit, dunning, MRR snapshot, cache health, weekly retro) Agent playbook
  • M1.3 Build `/api/admin-quality-trend` endpoint Learning-loop roadmap Phase 1
  • M1.4 Stripe webhook handler audit — confirm dunning fires on `invoice.payment_failed` implicit
  • M1.5 Run eval suite monthly going forward (cron) PR #63
  • M1.6 Watch first 30 days of Anthropic spend vs. cache hit rate (PR #56 instrumentation now real) PR #56
Low Low (16)
  • M2.1 Cache health dashboard (visual) PR #56 follow-up
  • M2.2 First prompt iteration based on real production data → MAIN_HINT_PROMPT_VERSION = 2 Learning-loop Phase 2
  • M2.3 Personalization layer (misconceptions injected into system prompt) Learning-loop Phase 2
  • M2.4 RAG layer (recent session summaries injected) Learning-loop Phase 2
  • M2.5 Active-learning review queue UI Learning-loop Phase 2
  • M2.6 Decide /faq vs /help consolidation strategy PR #65 audit S3
  • M2.7 Public roadmap page (`/roadmap`) Strengthen audit 10.1
  • M2.8 Public changelog page (`/changelog`) Strengthen audit 10.1
  • ST1 Trust center (`/trust` consolidating /privacy + /cookies + /child-safety + /sub-processors + security) Strengthen audit 9.3
  • ST2 A/B prompt testing in production via PostHog flags Learning-loop Phase 3
  • ST3 Per-grade-band prompt specialization Learning-loop Phase 3
  • ST4 DMCA agent registration + `/dmca` page PR #65 audit N3
  • ST5 Public sub-processor list page PR #65 audit N3
  • ST6 First hire (recruiting → onboarding → perf review) Agent playbook 8.x
  • ST7 Brand voice canon evolution after first 100 copy decisions Strengthen audit 10.3
  • ST8 Hintly Adults / Skill Coach spinoff `Hintly_Secondary_Products_Roadmap.md`
blocked Blocked (8)
  • LB2 Counsel sign-off on VPC method — checkbox-only consent may not satisfy COPPA "reasonably calculated" verification rule for a service that collects ong blocked PR #57 ANTHROPIC-COMPLIANCE-REVIEW + memory obs #13
  • LB3 Counsel sign-off on Google OAuth signup path — currently bypasses the COPPA consent checkboxes. Open question: are existing OAuth-signed-up parents validly cons blocked Memory obs #20 + PR #57
  • LB4 Privacy policy sub-processor table update — currently lists 6 of 9 actual sub-processors; missing Sentry, Upstash, Cloudflare. COPPA §312.4(d) violation if ship blocked PR #58 SUBPROCESSOR-DPA-MATRIX + PR #65 audit F6
  • LB5 Stripe dual-role disclosure — Stripe is both processor AND independent controller (fraud detection). Privacy policy doesn't capture this; counsel should co blocked PR #58 / PR #65 audit
  • LB6 Read the Supabase DPA PDF — binary PDF I couldn't extract; supabase holds 100% of user data including child profiles. blocked PR #58
  • LB7 Texas HB18 + state children's privacy laws review — partially enjoined but parts apply. Plus California CAADCA, NY Stop Addictive Feeds Act, etc. Founder + blocked Memory obs #4 / PR #57
  • LB8 DPA execution status check with each sub-processor — confirm executed DPA on file is the latest published version for: Anthropic, Supabase, Stripe, Vercel, Rese blocked PR #58
  • EV7 `EOOD_REGISTERED_ADDRESS` blocked **🔴 Launch-blocker** — CAN-SPAM postal address
Done (15)
  • EV1 `ANTHROPIC_API_KEY`
  • EV2 `SUPABASE_URL` + `SUPABASE_SERVICE_ROLE_KEY` + `VITE_SUPABASE_URL` + `VITE_SUPABASE_ANON_KEY`
  • EV3 `STRIPE_SECRET_KEY` + `STRIPE_WEBHOOK_SECRET` + extra-child price IDs
  • EV4 `RESEND_API_KEY`
  • EV5 `SITE_URL`
  • EV6 `CRON_SECRET`
  • EV8 `ADMIN_API_TOKEN`
  • EV9 `ALLOWED_DEMO_ORIGINS` (optional override)
  • EV10 `GEO_GATE_DISABLED`
  • EV11 `VITE_SENTRY_DSN` + `SENTRY_DSN`
  • EV12 `VITE_TURNSTILE_SITE_KEY` + `TURNSTILE_SECRET_KEY`
  • EV13 `UPSTASH_REDIS_REST_URL` + `UPSTASH_REDIS_REST_TOKEN`
  • EV14 `EVAL_BYPASS_KEY`
  • EV15 `ANALYTICS_API_KEY`
  • EV16 `DIGEST_RECIPIENT`

Milestones

  • 2026-05-15 · Compliance gate (DPAs + EOOD + sub-processor disclosure) (in 5d)
  • 2026-05-22 · Soft launch — named list of 30 (in 12d)
  • 2026-06-01 · Public launch (in 22d)
  • 2026-07-01 · First 100 paid signups (in 52d)
  • 2026-08-01 · PMF signal review (in 83d)

KPIs

KPITargetCurrentSource
Paid signups100 by 2026-07-010Stripe
Trial-to-paid>10%TBDPostHog
D7 retention>40%TBDPostHog
AI cost per active free user<€0.50/moTBDAnthropic billing

Spinoff scan

Inside
  • Hint-generation engine as standalone API — trigger: when 3rd-party request volume exists
  • Content catalog as licensable data set — trigger: when catalog reaches 1000+ vetted items
  • Pediatric Learning Insights ($5/mo upsell or $29 quarterly report) — uses session-pattern data — trigger: after Hintly hits $30K MRR · careful messaging (not amateur child psychology)
  • Reading Buddy ($5/mo add-on, Whisper STT + AI feedback) — bumps to Pro+ tier — trigger: after Hintly hits $30K MRR · ~4-week build
Outside
  • AI Worksheet Generator (worksheets.hintlyapp.com) — free → $5/mo · top-of-funnel · 60K+ monthly searches — trigger: Phase 1 in roadmap — after first 20 Hintly families onboarded
  • K-8 teachers + homeschool segment — trigger: when parent-side has 1k paying
  • Spanish edition (Hintly-ES) — leverages Petko Spanish — trigger: after PMF in English
  • Family Learning Letter (free email/funnel) — trigger: parked — fine as marketing via weekly summary; not standalone product
Spinoff
  • Hintly for Adults (Skill Coach) — $19/mo · same code, different prompt · could be $50K+ MRR Y2 — trigger: after Hintly hits $10K MRR direct-to-consumer · open ProjectB folder · Petko HIGH interest
  • Hintly for Schools (B2B white-label) — $5-15/student/year · enterprise sales — trigger: parked until $50K MRR direct-to-consumer · 3-6mo sales cycles + FERPA burden
  • Kid-safe AI certification service — trigger: after 12mo of safety-layer learnings

P&L scenarios (7)

Pricing: Pro $14.99/mo ($9.92 annual) · Family Complete $29.99/mo ($20.75 annual) · Free tier (5 sessions/wk)
LineM1 launch (10 paid)M3 traction (50)M6 growth (200)M12 Y1 (500)Y2 (2,000)Y3 (5,000)Scale milestone (10k)
Revenue/mo$170$850$3,400$8,500$34,000$85,000$170,000
− Variable−$20−$100−$400−$1,000−$4,000−$10,000−$20,000
− Fixed (€156)−$156−$156−$156−$156−$156−$156−$156
OI/mo-$6$594$2,844$7,344$29,844$74,844$149,844
Annual-$72/yr$7,128/yr$34,128/yr$88,128/yr$358,128/yr$898,128/yr$1,798,128/yr
5 canonical docs
  • Hintly/CLAUDE.md
  • Hintly/Hintly_Internal_Agent_Scope.md
  • Hintly/LAUNCH_AUDIT.md
  • Hintly/LAUNCH_DASHBOARD_CHECKS.md
  • Founder Operating/FOUNDER-AI-PLAYBOOK.md

Spreadsheet OS collection

9 generated spreadsheet products (Airbnb, Real Estate, Wedding, Renovation, Grocery, Debt, Budget, Tax, Invoice).

Building P2 · this month
Stage
artifacts exist as .xlsx + build_*.py scripts on Desktop
Market
solo operators / households / niche use cases
Model
One-time digital product. Etsy / Gumroad inventory.
Path
. (root Desktop — see build_*.py scripts)
Last update
pre-2026-05-09
Next action
Pick top 2-3 by demand evidence. Write a sales page. List on Etsy. Treat the rest as portfolio depth.
Biggest risk
9 SKUs with no listing is 0 listings. Pick a winner.

Open tasks (cross-portfolio: 1)

  • Medium Business List at least one Spreadsheet OS on Etsy T-004 due 36d

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

Tax Project

Personal tax tracking. README + archive.

Building P2 · this month
Stage
docs only
Market
self-use
Model
not applicable (operational)
Path
Tax Project
Last update
pre-2026-05-09
Next action
Reconcile with Tax & Invoice Tracker.xlsx + Quarterly_Tax_Estimator_2026.xlsx. Single source of truth.
Biggest risk
Three places tracking the same numbers (project + 2 spreadsheets).

Open tasks (cross-portfolio: 1)

  • High Financial Reconcile tax tracking — pick one source of truth T-002 due 21d

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

1 canonical doc
  • Tax Project/README.md

Freelancer OS / SoloistOS

Productized OS template for solo creators. Multiple versions (v4-v6).

Building P3 · this quarter
Stage
v6 latest, marketing assets exist, Etsy shop setup guide drafted
Market
Solo creators / freelancers
Model
One-time digital product (Etsy / Gumroad / own site)
Path
Freelancer OS
Last update
pre-2026-05-09
Next action
Decide packaging and channel. List on Etsy or Gumroad. Stop versioning, start selling.
Biggest risk
Versioning forever without launching. v4 → v5 → v6 with no v1 in market.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

2 canonical docs
  • Freelancer OS/Freelancer OS — Project Summary.md
  • Freelancer OS/etsy-shop-setup-guide.html

MDM Engine (Petko Digital Products Ideas)

Webapp + sample data + legal + marketing. Has the Founder Playbook in-folder.

Building P3 · this quarter
Stage
webapp + legal + marketing scaffolded
Market
TBD
Model
TBD
Path
Petko Digital Products Ideas
Last update
pre-2026-05-09
Next action
Read README to recover context. Likely needs a focus decision vs Pet+Owner OS.
Biggest risk
Overlap with Pet+Owner OS unclear.

FOUNDER-TASKS.md (43 open)

Blocker Blocker (7)
  • LB1 Complete EOOD incorporation Cannot accept payment without legal entity
  • LB2 Switch Stripe to live mode Needed for actual revenue
  • LB3 Fill `[BRACKETS]` in legal pages — entity name, address, governing law Privacy/Terms/DPA reference placeholder values today (verifi
  • LB4 ~~Verify RLS isolates tenants end-to-end~~ ~~A bug here = cross-tenant data leak~~
  • LB5 ~~Verify `/api/account/delete` purges blobs + Stripe customer~~ ~~GDPR Art. 17 risk~~
  • LB6 ~~Add `.github/workflows/ci.yml` (typecheck + vitest)~~ ~~Regressions in engine ship undetected today~~
  • LB7 Penetration test or independent security review Cheap insurance before money flows
High High (15)
  • PL1 Define ICP in 2-3 sentences and rewrite landing hero Today: "distributors / retailers / brand owners" t
  • PL2 Create `docs/BRAND-VOICE.md` Voice drifts across landing/emails/pricing pages today
  • PL3 Create `.env.example` documenting all 12 env vars Currently nothing documents required env (verified by grep)
  • PL4 Add Sentry (error tracking) `error.tsx:18` literally says "we'd ship this to S
  • PL5 Add status page (UptimeRobot or Better Stack free tier) Trust signal in sales pitch
  • PL6 Wire Healthchecks.io pings to both crons `/api/cron/purge-blobs` + `/api/cron/send-emails` are curren
  • PL7 Record 2-minute Loom of upload→clean→download Embed on landing above fold
  • PL8 Send first 5 cold outreach messages from `marketing/sales-pitch.md` Lowest-CAC distribution test
  • PL9 Send first 3 partnership outreach (SAP migration consultancies) Highest-leverage channel
  • PL10 Pricing tiers redesign (Free / Pro / Team / Enterprise with row caps) Today $149 flat undermonetizes large customers
  • PL11 Cookie consent banner EU regulatory hygiene
  • PL12 Resend account + verify sending domain Onboarding emails are coded but inert until this is set up —
  • PL13 Configure Clerk webhook with signing secret + set `CLERK_WEBHOOK_SECRET` Triggers the welcome email
  • PL14 Set `CRON_SECRET` env var for hourly cron auth Prevents external abuse of `/api/cron/send-emails`
  • PL15 Audit `npm audit` findings — 7 vulnerabilities (6 moderate, 1 high) Surfaced when CI first ran. Some may need force-fixes that b
Medium Medium (7)
  • M1.1 PostHog event instrumentation — funnel events (signup, first run, first download, upgrade) Pending
  • M1.2 Add Playwright smoke suite with render-error assertions (per playbook Part 7) Pending
  • M1.3 Add Lighthouse-CI to CI workflow Pending
  • M1.4 Knowledge base / FAQ page on landing Pending
  • M1.5 Onboarding email sequence verified working end-to-end (after PL12-14) Pending
  • M1.6 Customer-success motion: NPS survey at run 5; 30-day check-in template Pending
  • M1.7 First case study + reference customer permission Pending
Low Low (14)
  • M23.1 Team / multi-user accounts (Clerk Organizations) Build when first customer asks
  • M23.2 Public REST API + API keys Build when first automation customer asks
  • M23.3 Persistent connectors (S3, SFTP, vendor email inbox) Per-request
  • M23.4 Run comparison view (diff two runs of the same vendor) Pending
  • M23.5 Tier 2 custom output (full YAML profile upload) Build when 3+ customers ask
  • M23.6 DPP JSON-LD output writer (currently emits CSV; ESPR submission needs JSON-LD) Pending
  • M23.7 Audit log of admin actions Build when first regulated-industry customer surfaces
  • M23.8 SSO (Clerk Enterprise tier) Build when first enterprise prospect blocks on it
  • S1 Channel partnerships — pick 3 SAP migration consultancies, formal partner program Per E2E audit: highest-leverage distribution lever, not yet
  • S2 Pricing redesign with usage-based + per-user tiers Per E2E audit: today's flat $149 leaves 30-60% on the t
  • S3 DPP regulatory deep-dive — full ESPR field set, JSON-LD output, SCIP integration Wedge bet: regulatory tailwind 2026-2027
  • S4 SAP-deeper — more SAP profiles (CC1, MDG/IDOC/iDoc, S/4HANA Migration Cockpit XML) Wedge bet: 2027 S/4HANA migration deadline
  • S5 Schema-broader — let customers define their own schemas via UI Wedge bet: opens horizontal markets (medical devices, food,
  • S6 SOC 2 / ISO 27001 6-12 month process; only worth starting when 1+ enterprise p

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

2 canonical docs
  • Petko Digital Products Ideas/FOUNDER-AI-PLAYBOOK.md
  • Petko Digital Products Ideas/README.md

PetkoOS — Execution Engine ("Finish What Matters")

SaaS enforcing execution discipline (max 2 active, completion protocol, scoring, streaks). Deferred 2026-05-10 per D-005 — code preserved, awaiting trigger.

Building P3 · this quarter
Stage
Far more built than registry previously implied. 14 Supabase migrations (profiles, tasks, execution sessions, closure checklists, scores, WIP triggers, daily scores, streaks, inbox status, hourly rate, due date, recurring tasks, Stripe subscriptions, AI usage caps). Next.js app with auth + dashboard + execute + API routes. Components for analytics, dashboard, execution, onboarding, quick-capture, review, settings, tasks.
Market
Freelancers, tech professionals, solo operators, builders
Model
SaaS — Free / Pro €8-15/mo / Premium €20-40/mo / Team later
Path
PetkoOS
Last update
2026-05-10
Next action
No active work until trigger fires. Re-evaluate at 2026-06-01 (Hintly launch) and 2026-08-01 (Hintly stability gate).
Biggest risk
Code dormancy past 6 months (~2026-12) — revival cost grows; possibly switch to permanent shelve at that point.
Revival trigger
Activate as P2 commit IF (a) ClinicFlow validation fails by 2026-05-31 (PetkoOS replaces failed ClinicFlow), OR (b) Hintly reaches 100 paid users + <8% monthly churn by 2026-08-01 (Hintly stable enough to share founder bandwidth).

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

4 canonical docs
  • PetkoOS/Execution & Decision OS.txt
  • PetkoOS/execution-os/README.md
  • PetkoOS/execution-os/CLAUDE.md
  • PetkoOS/execution-os/AGENTS.md

AI Agent (ClinicFlow)

AI receptionist that responds, qualifies, and books patients 24/7.

B2BAIclinicSpainBGvalidation
Validating P2 · this month
Stage
PRD complete, pre-MVP
Market
Clinics (dental, aesthetic, physio) — Bulgaria-first, EU expansion
Model
Setup fee + €29-99/mo SaaS
Path
AI Agent
Last update
pre-2026-05-09
Next action
Identify 20 BG clinics, run outreach, book 2-3 demos. Build chatbot + Make flow only after first demo.
Biggest risk
Building before validating willingness to pay. The PRD assumes the customer exists.

Open tasks (cross-portfolio: 1)

  • Medium Business ClinicFlow validation — outreach to 20 BG clinics T-006 due 21d

Milestones

  • 2026-05-31 · 20 BG clinics outreach complete (in 21d)
  • 2026-06-15 · 2-3 demos booked (in 36d)
  • 2026-07-01 · First MVP delivered manually to 1 clinic (in 52d)
  • 2026-09-01 · Spain validation cohort begins (Spanish-language) (in 114d)

KPIs

KPITargetCurrentSource
Discovery calls completed20 by 2026-06-300manual log
Qualified demos booked5 by 2026-06-300Calendly
First paying clinic1 by 2026-08-010Stripe

Spinoff scan

Inside
  • Appointment booking as separate AI receptionist SKU — trigger: after dental fit confirmed
  • Lead-qualification engine as standalone tool — trigger: if non-clinic verticals show demand
Outside
  • Aesthetic, physio, vet, GP — trigger: after dental MVP delivers ROI
  • Spain expansion (Spanish — Petko native) — trigger: after BG fit confirmed
  • Mexico / LatAm Spanish markets — trigger: after Spain proves Spanish demand
Spinoff
  • Clinic-data ontology as industry-standard data model (MDM connection) — trigger: after 50+ clinic implementations

P&L scenarios (4)

Pricing: €500-1500 setup + €29-99/mo SaaS (per AI Agent PRD)
LineM3 first paying clinicM6 5 clinicsM12 20 clinicsSpain expansion (M18+)
Revenue/mo€99€495€2,000€5,000
− Variable−€15−€75−€300−€800
− Fixed (€76)−€76−€76−€76−€76
OI/mo€8€344€1,624€4,124
Annual€96/yr€4,128/yr€19,488/yr€49,488/yr
1 canonical doc
  • AI Agent/AI AGENT.txt

ProjectE

Has validation folder + GENERAL MEMORY.

Validating P3 · this quarter
Stage
validation in progress
Market
unknown
Model
unknown
Path
ProjectE
Last update
pre-2026-05-09
Next action
Read validation/ to recover context. Rename to real product name.
Biggest risk
Same.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

AI Help Assistant (productized service)

5-day productized service installing AI help widget on B2B SaaS. $2,500 flat. Deferred 2026-05-10 per D-004 — product-only portfolio strategy.

Concept P4 · someday
Stage
brief + offer page + outreach playbook + tech spec all drafted; not selling. Material preserved for trigger-fire reactivation.
Market
B2B SaaS founders, $10K-500K MRR, US/UK/EU/CA/AU
Model
Productized service — $2,500/delivery, ~5 days each, target 4/quarter = €40K/quarter
Path
ProjectA
Last update
2026-05-10
Next action
No active work until trigger fires. Material preserved.
Biggest risk
Activation without strategic re-think = founder time fragmentation. The defer is the lower-risk path.
Activation trigger
Reactivate IF (a) Ingram Micro income disrupted (layoff, role change), OR (b) Hintly fails to clear €1K MRR by 2026-09-01, OR (c) Petko explicitly decides services bridge is wanted (force re-decision). Reactivation should consider MERGING ProjectA + ClinicFlow into a single B2B AI-deflection product (per spinoff scan refresh), NOT activating as a separate third commit slot.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L scenarios (3)

Pricing: $2,500 flat per delivery (USD); ~€2,300 EUR
LineQ1 baseline (1 delivery)Q1 target (4 deliveries)Q1 stretch (8 deliveries)
Revenue/mo€766€3,066€6,133
− Variable−€100−€400−€800
− Fixed (€0)−€0−€0−€0
OI/mo€666€2,666€5,333
Annual€7,992/yr€31,992/yr€63,996/yr
4 canonical docs
  • ProjectA/brief.md
  • ProjectA/offer.md
  • ProjectA/outreach-playbook.md
  • ProjectA/tech-spec.md

Lookup Engine

Concept (only GENERAL MEMORY in directory).

Concept P4 · someday
Stage
empty
Market
unknown
Model
unknown
Path
Lookup Engine
Last update
pre-2026-05-09
Next action
Document the idea or shelve.
Biggest risk
Same as Avero — drift.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

Pet+Owner Operating System

Digital pet identity + lifecycle platform (consumer + vet B2B). PRD only, no live build. Reclassified concept 2026-05-10 per T-008 analysis.

Concept P4 · someday
Stage
PRD + Turbo monorepo scaffolding (no real build evident)
Market
Pet owners (primary), vets (B2B), groomers/boarders
Model
TBD — likely freemium for owners + B2B seat for vets
Path
Pet+OwnerOperatingSystem
Last update
2026-05-10
Next action
Decision needed — fold into T-001 (which 2 of 7+ concepts to keep alive). Currently no active investment.
Biggest risk
Confused with MDM Engine (it isn't — see T-008 analysis). Real risk: occupies a concept-priority slot that hasn't moved in 6+ months.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

1 canonical doc
  • Pet+OwnerOperatingSystem/Pet OS.txt

Avero — "Dinner, decided."

Recipe recommendation engine MVP (50 recipes, mobile-first React). Day 0 triaged 2026-05-09.

Shelved P4 · someday
Stage
single .jsx (structured-MVP React pattern)
Market
consumer (recipe / dinner-decision)
Model
TBD — but space is crowded
Path
Avero
Last update
2026-05-09
Next action
Biggest risk
Shelved reason
Consumer recipe space too crowded (Yummly, Tasty, Mealime, Whisk all free). No clear commercial path. Code retained as React structured-MVP reference.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

1 canonical doc
  • Avero/avero (1).jsx

DreamQuest — Bedtime Adventures

AI-generated personalized bedtime stories for kids. Shelved 2026-05-10 per D-006.

Shelved P4 · someday
Stage
prototype HTMLs + PRD + action items + app build artifacts
Market
Parents of 3-9 year olds
Model
TBD — likely freemium with paid story packs / subscription
Path
DreamQuest
Last update
2026-05-10
Next action
Biggest risk
Shelved reason
Honest read — Petko isn't building this in next 6 months. Hintly + ClinicFlow (with deferred PetkoOS, deferred ProjectA) saturate commit slots. Killing outright loses real upside (DreamQuest as Hintly upsell to same parent customer).
Revival trigger
Activate IF Hintly hits BOTH (a) 1,000+ paid families AND (b) 30%+ of churned parents cite "kid lost interest in homework focus" or similar bonding/engagement gap that DreamQuest's emotional-bonding angle directly addresses. Reverse triggers: storyteller co-founder approach, organic parent pull for bedtime features, competitive forcing function.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

3 canonical docs
  • DreamQuest/DreamQuest PRD.txt
  • DreamQuest/DreamQuest action items.txt
  • DreamQuest/DreamQuest notes.txt

LinguaLeap

AI conversation practice for 7 languages (incl. Spanish + Greek — plays to Petko's Spanish asset). Day 0 triaged 2026-05-09.

Shelved P4 · someday
Stage
prototype JSX + page + API route
Market
language learners (broadly)
Model
TBD — space is brutally crowded (Duolingo, Babbel, ChatGPT)
Path
LinguaLeap
Last update
2026-05-09
Next action
Biggest risk
Shelved reason
Language learning crowded with free + dominant players. Petko's Spanish + Bulgarian context could find a niche, but not without explicit positioning work.
Revival trigger
EITHER (a) a niche emerges leveraging Petko's Spanish + Bulgarian profile (e.g., "Spanish for BG professionals", "BG → Spain relocation"), OR (b) Hintly success creates adjacent kid-language angle.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

3 canonical docs
  • LinguaLeap/page.jsx
  • LinguaLeap/lingualeap.jsx
  • LinguaLeap/api-conversation-route.js

ProjectB

Empty placeholder folder, no documented intent. SHELVED 2026-05-09 triage.

Shelved P4 · someday
Stage
empty
Market
n/a
Model
n/a
Path
ProjectB
Last update
2026-05-09
Next action
Biggest risk
Shelved reason
No documented intent, no recoverable context. See inbox/killed-ideas.md.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

ProjectC

Empty placeholder + Founder Playbook copy, no documented intent. SHELVED 2026-05-09 triage.

Shelved P4 · someday
Stage
empty
Market
n/a
Model
n/a
Path
ProjectC
Last update
2026-05-09
Next action
Biggest risk
Shelved reason
No documented intent, no recoverable context. Playbook copy without project. See inbox/killed-ideas.md.

Milestones

No milestones set. Add to registry/projects.yaml.

KPIs

No KPIs set. Add to registry/projects.yaml.

Spinoff scan

No spinoff scan. Run agents/spinoff-strategist.md.

P&L

No financials defined. Add a financials: block to registry/projects.yaml for this project.

1 canonical doc
  • ProjectC/FOUNDER-AI-PLAYBOOK.md

Tasks

Cross-portfolio tasks (7)

Founder-owned tasks that span projects (registry/tasks.yaml). Engineering-internal tasks live in each project's repo (FOUNDER-TASKS.md table below).

Severity:
Type:
Project:
SeverityTypeTaskNext stepDue
BlockerComplianceHintly compliance gate — DPAs + EOOD + sub-processor disclosure doing
T-005 · hintly
(a) Email accountant re: EOOD registered office address (use accountant's address as c/o to unblock launch + permanent registered office). (b) Schedule counsel intake call (LB11) — master gate for 7 LB items. (c) Knock out Common Sense Privacy submission (LB13, Petko-only, no dependencies). 2026-05-15 (in 5d)
HighFinancialReconcile tax tracking — pick one source of truth
T-002 · tax-project
30 minutes this week. Start by reading Tax Project/README.md.2026-05-31 (in 21d)
HighStrategicDecide which 2 of the 6+ concept projects to keep alive
T-001 · portfolio
Quarterly C-suite re-run on portfolio — Q3 2026 cycle.2026-06-30 (in 51d)
MediumBusinessClinicFlow validation — outreach to 20 BG clinics
T-006 · ai-agent-clinicflow
90-min block. Build clinic list, draft cold message, send.2026-05-31 (in 21d)
MediumBusinessList at least one Spreadsheet OS on Etsy
T-004 · spreadsheet-os-collection
Pick the highest-evidence-of-demand SKU. Real Estate Deal OS is plausible.2026-06-15 (in 36d)
LowOperationalCowork plugins — package Personal OS workflows after 2 weeks of basic Cowork use
T-014 · portfolio
⛔ blocked by T-013
At 2026-05-31 Friday review, list the workflows invoked >3 times in past 2 weeks. Package those as Cowork plugins per mobile/cowork-setup.md.2026-05-31 (in 21d)
LowOperationalSet up project memory directories for every shipping/building project
T-009 · portfolio
Re-run python automation/check_drift.py at next Friday curation.no due

Per-project FOUNDER-TASKS.md (99 open across 2 projects)

Auto-aggregated from each project's docs/FOUNDER-TASKS.md (and worktree fallback). To complete an item: edit the source file, then re-run python automation/build_dashboard.py.

Severity:
Project:
SeverityItemSource
BlockerSet `EOOD_REGISTERED_ADDRESS` in Vercel env — every transactional email currently shows a placeholder. CAN-SPAM Act §316.4(a)(5) requires valid postal address.
LB1 · hintly · Compliance & legal
PR #66 audit pass 2 / `_emailLayout.js`
blockedCounsel sign-off on VPC method — checkbox-only consent may not satisfy COPPA "reasonably calculated" verification rule for a service that collects ongoing learning records. May need stronger blocked
LB2 · hintly · Compliance & legal
PR #57 ANTHROPIC-COMPLIANCE-REVIEW + memory obs #13
blockedCounsel sign-off on Google OAuth signup path — currently bypasses the COPPA consent checkboxes. Open question: are existing OAuth-signed-up parents validly consented? blocked
LB3 · hintly · Compliance & legal
Memory obs #20 + PR #57
blockedPrivacy policy sub-processor table update — currently lists 6 of 9 actual sub-processors; missing Sentry, Upstash, Cloudflare. COPPA §312.4(d) violation if shipped as-is. blocked
LB4 · hintly · Compliance & legal
PR #58 SUBPROCESSOR-DPA-MATRIX + PR #65 audit F6
blockedStripe dual-role disclosure — Stripe is both processor AND independent controller (fraud detection). Privacy policy doesn't capture this; counsel should confirm wording. blocked
LB5 · hintly · Compliance & legal
PR #58 / PR #65 audit
blockedRead the Supabase DPA PDF — binary PDF I couldn't extract; supabase holds 100% of user data including child profiles. blocked
LB6 · hintly · Compliance & legal
PR #58
blockedTexas HB18 + state children's privacy laws review — partially enjoined but parts apply. Plus California CAADCA, NY Stop Addictive Feeds Act, etc. Founder + counsel decide which states are open. blocked
LB7 · hintly · Compliance & legal
Memory obs #4 / PR #57
blockedDPA execution status check with each sub-processor — confirm executed DPA on file is the latest published version for: Anthropic, Supabase, Stripe, Vercel, Resend, Sentry, PostHog, Upstash, Cloudflare blocked
LB8 · hintly · Compliance & legal
PR #58
BlockerEOOD formation finalized
LB9 · hintly · Operational launch-blockers
Memory `project_launch_plan_2026_05.md` B-list
BlockerLiability insurance
LB10 · hintly · Operational launch-blockers
Memory B-list
BlockerCounsel retention
LB11 · hintly · Operational launch-blockers
Memory B-list
BlockerVerify Vercel env vars all set — full list below in 🟠 section. `ANTHROPIC_API_KEY`, `SUPABASE_*`, `STRIPE_*`, `RESEND_API_KEY`, `EOOD_REGISTERED_ADDRESS`, etc.
LB12 · hintly · Operational launch-blockers
PR #66 + .env.example
BlockerCommon Sense Privacy submission
LB13 · hintly · Operational launch-blockers
Memory B-list / `docs/compliance/common-sense-privacy-submission.md`
blocked`EOOD_REGISTERED_ADDRESS` blocked
EV7 · hintly · Vercel environment variables
**🔴 Launch-blocker** — CAN-SPAM postal address
HighConfirm `support@hintlyapp.com` routes to founder inbox
EM1 · hintly · Email forwarders + DNS
always
HighConfirm `team@hintlyapp.com` routes (referenced in /careers)
EM2 · hintly · Email forwarders + DNS
PR #65 audit S6
HighConfirm `privacy@hintlyapp.com` routes (referenced in privacy + KB)
EM3 · hintly · Email forwarders + DNS
PR #65 audit S6
HighSPF / DKIM / DMARC for hintlyapp.com → Resend (deliverability)
EM4 · hintly · Email forwarders + DNS
always
HighTest welcome email lands in Gmail / Outlook / iCloud inbox (not spam)
EM5 · hintly · Email forwarders + DNS
implicit
HighClaim `@hintlyapp` on X (or remove footer icon)
SO1 · hintly · Social handles
PR #65 audit S7
HighClaim `@hintlyapp` on Instagram (or remove icon)
SO2 · hintly · Social handles
PR #65 audit S7
HighClaim `@hintlyapp` on LinkedIn company page (or remove icon)
SO3 · hintly · Social handles
PR #65 audit S7
HighFirst eval-suite run against launch baseline → commit `eval/quality-suite/results/baseline-2026-05.md`
OP1 · hintly · Beta / launch operations
PR #59 STRESS-TEST-PLAN + PR #63 LEARNING-LOOP
HighFirst k6 stress-test run against a Vercel preview → commit `docs/STRESS-TEST-RESULTS.md`
OP2 · hintly · Beta / launch operations
PR #59
HighLighthouse-CI baseline numbers committed (after Lighthouse runs in PR #47)
OP3 · hintly · Beta / launch operations
PR #47
HighHealthchecks.io account → set `HEALTHCHECKS_PING_URL` env var
OP4 · hintly · Beta / launch operations
Memory B-list / Agent playbook 5.10
HighSlack workspace incoming webhook → set `SLACK_ALERT_WEBHOOK_URL` env var
OP5 · hintly · Beta / launch operations
Agent playbook Phase 1
HighPostHog read-only API key → set `POSTHOG_API_KEY` for daily founder digest agent
OP6 · hintly · Beta / launch operations
Agent playbook 9.2
HighPostHog feature-flag config (`signup-trial-variant`, etc.) reviewed and shipped
OP7 · hintly · Beta / launch operations
Memory B-list
HighDecide founder-digest email destination + send time (default 7am EST to founder)
OP8 · hintly · Beta / launch operations
Agent playbook 9.2
HighGSC (Google Search Console) validate-fix on any pre-launch crawl errors
OP9 · hintly · Beta / launch operations
Memory B-list
HighVerify Supabase `pgcron` jobs scheduled correctly (not skipping)
OP10 · hintly · Beta / launch operations
implicit
HighTest the full signup → onboarding → first session → weekly summary flow end-to-end with a clean test account
OP11 · hintly · Beta / launch operations
implicit
HighSend a test welcome email + weekly digest to yourself; confirm no placeholder text
OP12 · hintly · Beta / launch operations
PR #66 / this PR
MediumReal testimonials collection → replace placeholder spots with quotes
M1.1 · hintly · 🟡 Post-launch month 1
Memory `project_overhaul_plan.md`
MediumWire 8 Phase-1 agents from agent playbook (incident response, daily digest, error budget, COPPA audit, dunning, MRR snapshot, cache health, weekly retro)
M1.2 · hintly · 🟡 Post-launch month 1
Agent playbook
MediumBuild `/api/admin-quality-trend` endpoint
M1.3 · hintly · 🟡 Post-launch month 1
Learning-loop roadmap Phase 1
MediumStripe webhook handler audit — confirm dunning fires on `invoice.payment_failed`
M1.4 · hintly · 🟡 Post-launch month 1
implicit
MediumRun eval suite monthly going forward (cron)
M1.5 · hintly · 🟡 Post-launch month 1
PR #63
MediumWatch first 30 days of Anthropic spend vs. cache hit rate (PR #56 instrumentation now real)
M1.6 · hintly · 🟡 Post-launch month 1
PR #56
LowCache health dashboard (visual)
M2.1 · hintly · 🟢 Post-launch month 2-3
PR #56 follow-up
LowFirst prompt iteration based on real production data → MAIN_HINT_PROMPT_VERSION = 2
M2.2 · hintly · 🟢 Post-launch month 2-3
Learning-loop Phase 2
LowPersonalization layer (misconceptions injected into system prompt)
M2.3 · hintly · 🟢 Post-launch month 2-3
Learning-loop Phase 2
LowRAG layer (recent session summaries injected)
M2.4 · hintly · 🟢 Post-launch month 2-3
Learning-loop Phase 2
LowActive-learning review queue UI
M2.5 · hintly · 🟢 Post-launch month 2-3
Learning-loop Phase 2
LowDecide /faq vs /help consolidation strategy
M2.6 · hintly · 🟢 Post-launch month 2-3
PR #65 audit S3
LowPublic roadmap page (`/roadmap`)
M2.7 · hintly · 🟢 Post-launch month 2-3
Strengthen audit 10.1
LowPublic changelog page (`/changelog`)
M2.8 · hintly · 🟢 Post-launch month 2-3
Strengthen audit 10.1
LowTrust center (`/trust` consolidating /privacy + /cookies + /child-safety + /sub-processors + security)
ST1 · hintly · ⚫ Strategic / discretionary
Strengthen audit 9.3
LowA/B prompt testing in production via PostHog flags
ST2 · hintly · ⚫ Strategic / discretionary
Learning-loop Phase 3
LowPer-grade-band prompt specialization
ST3 · hintly · ⚫ Strategic / discretionary
Learning-loop Phase 3
LowDMCA agent registration + `/dmca` page
ST4 · hintly · ⚫ Strategic / discretionary
PR #65 audit N3
LowPublic sub-processor list page
ST5 · hintly · ⚫ Strategic / discretionary
PR #65 audit N3
LowFirst hire (recruiting → onboarding → perf review)
ST6 · hintly · ⚫ Strategic / discretionary
Agent playbook 8.x
LowBrand voice canon evolution after first 100 copy decisions
ST7 · hintly · ⚫ Strategic / discretionary
Strengthen audit 10.3
LowHintly Adults / Skill Coach spinoff
ST8 · hintly · ⚫ Strategic / discretionary
`Hintly_Secondary_Products_Roadmap.md`
BlockerComplete EOOD incorporation
LB1 · mdm-engine · 🔴 Launch-blockers
Cannot accept payment without legal entity
BlockerSwitch Stripe to live mode
LB2 · mdm-engine · 🔴 Launch-blockers
Needed for actual revenue
BlockerFill `[BRACKETS]` in legal pages — entity name, address, governing law
LB3 · mdm-engine · 🔴 Launch-blockers
Privacy/Terms/DPA reference placeholder values today (verified `webapp/src/lib/l
Blocker~~Verify RLS isolates tenants end-to-end~~
LB4 · mdm-engine · 🔴 Launch-blockers
~~A bug here = cross-tenant data leak~~
Blocker~~Verify `/api/account/delete` purges blobs + Stripe customer~~
LB5 · mdm-engine · 🔴 Launch-blockers
~~GDPR Art. 17 risk~~
Blocker~~Add `.github/workflows/ci.yml` (typecheck + vitest)~~
LB6 · mdm-engine · 🔴 Launch-blockers
~~Regressions in engine ship undetected today~~
BlockerPenetration test or independent security review
LB7 · mdm-engine · 🔴 Launch-blockers
Cheap insurance before money flows
HighDefine ICP in 2-3 sentences and rewrite landing hero
PL1 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Today: "distributors / retailers / brand owners" too broad for outboun
HighCreate `docs/BRAND-VOICE.md`
PL2 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Voice drifts across landing/emails/pricing pages today
HighCreate `.env.example` documenting all 12 env vars
PL3 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Currently nothing documents required env (verified by grep)
HighAdd Sentry (error tracking)
PL4 · mdm-engine · 🟠 Pre-launch (before public marketing push)
`error.tsx:18` literally says "we'd ship this to Sentry... for now con
HighAdd status page (UptimeRobot or Better Stack free tier)
PL5 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Trust signal in sales pitch
HighWire Healthchecks.io pings to both crons
PL6 · mdm-engine · 🟠 Pre-launch (before public marketing push)
`/api/cron/purge-blobs` + `/api/cron/send-emails` are currently silent
HighRecord 2-minute Loom of upload→clean→download
PL7 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Embed on landing above fold
HighSend first 5 cold outreach messages from `marketing/sales-pitch.md`
PL8 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Lowest-CAC distribution test
HighSend first 3 partnership outreach (SAP migration consultancies)
PL9 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Highest-leverage channel
HighPricing tiers redesign (Free / Pro / Team / Enterprise with row caps)
PL10 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Today $149 flat undermonetizes large customers
HighCookie consent banner
PL11 · mdm-engine · 🟠 Pre-launch (before public marketing push)
EU regulatory hygiene
HighResend account + verify sending domain
PL12 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Onboarding emails are coded but inert until this is set up — see `webapp/docs/on
HighConfigure Clerk webhook with signing secret + set `CLERK_WEBHOOK_SECRET`
PL13 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Triggers the welcome email
HighSet `CRON_SECRET` env var for hourly cron auth
PL14 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Prevents external abuse of `/api/cron/send-emails`
HighAudit `npm audit` findings — 7 vulnerabilities (6 moderate, 1 high)
PL15 · mdm-engine · 🟠 Pre-launch (before public marketing push)
Surfaced when CI first ran. Some may need force-fixes that break versions; needs
MediumPostHog event instrumentation — funnel events (signup, first run, first download, upgrade)
M1.1 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumAdd Playwright smoke suite with render-error assertions (per playbook Part 7)
M1.2 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumAdd Lighthouse-CI to CI workflow
M1.3 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumKnowledge base / FAQ page on landing
M1.4 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumOnboarding email sequence verified working end-to-end (after PL12-14)
M1.5 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumCustomer-success motion: NPS survey at run 5; 30-day check-in template
M1.6 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
MediumFirst case study + reference customer permission
M1.7 · mdm-engine · 🟡 Month 1 (after customer #1 lands)
Pending
LowTeam / multi-user accounts (Clerk Organizations)
M23.1 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Build when first customer asks
LowPublic REST API + API keys
M23.2 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Build when first automation customer asks
LowPersistent connectors (S3, SFTP, vendor email inbox)
M23.3 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Per-request
LowRun comparison view (diff two runs of the same vendor)
M23.4 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Pending
LowTier 2 custom output (full YAML profile upload)
M23.5 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Build when 3+ customers ask
LowDPP JSON-LD output writer (currently emits CSV; ESPR submission needs JSON-LD)
M23.6 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Pending
LowAudit log of admin actions
M23.7 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Build when first regulated-industry customer surfaces
LowSSO (Clerk Enterprise tier)
M23.8 · mdm-engine · 🟢 Month 2-3 (post initial product-market signal)
Build when first enterprise prospect blocks on it
LowChannel partnerships — pick 3 SAP migration consultancies, formal partner program
S1 · mdm-engine · ⚫ Strategic / wedge bets
Per E2E audit: highest-leverage distribution lever, not yet attempted
LowPricing redesign with usage-based + per-user tiers
S2 · mdm-engine · ⚫ Strategic / wedge bets
Per E2E audit: today's flat $149 leaves 30-60% on the table
LowDPP regulatory deep-dive — full ESPR field set, JSON-LD output, SCIP integration
S3 · mdm-engine · ⚫ Strategic / wedge bets
Wedge bet: regulatory tailwind 2026-2027
LowSAP-deeper — more SAP profiles (CC1, MDG/IDOC/iDoc, S/4HANA Migration Cockpit XML)
S4 · mdm-engine · ⚫ Strategic / wedge bets
Wedge bet: 2027 S/4HANA migration deadline
LowSchema-broader — let customers define their own schemas via UI
S5 · mdm-engine · ⚫ Strategic / wedge bets
Wedge bet: opens horizontal markets (medical devices, food, chemicals)
LowSOC 2 / ISO 27001
S6 · mdm-engine · ⚫ Strategic / wedge bets
6-12 month process; only worth starting when 1+ enterprise prospect blocks

Kanban — this week

Unified across cross-portfolio tasks (registry/tasks.yaml) AND per-project FOUNDER-TASKS.md blockers + high-severity items · +51 medium/low FOUNDER-TASKS items rolled up (see Tasks tab for full list). To move a registry task to "doing", add kanban: doing to its yaml. To complete: python automation/complete_task.py <id> [project]. Max 2 in DOING (per playbooks/11-execution-discipline.md).

TODO

54
Blocker FOUNDER-TASKS LB1
Set `EOOD_REGISTERED_ADDRESS` in Vercel env — every transactional email currently shows a placeholder. CAN-SPAM Act §316.4(a)(5) requires valid postal address.
hintly · Compliance & legal
Blocker FOUNDER-TASKS LB9
EOOD formation finalized
hintly · Operational launch-blockers
Blocker FOUNDER-TASKS LB10
Liability insurance
hintly · Operational launch-blockers
Blocker FOUNDER-TASKS LB11
Counsel retention
hintly · Operational launch-blockers
Blocker FOUNDER-TASKS LB12
Verify Vercel env vars all set — full list below in 🟠 section. `ANTHROPIC_API_KEY`, `SUPABASE_*`, `STRIPE_*`, `RESEND_API_KEY`, `EOOD_REGISTERED_ADDRESS`, etc.
hintly · Operational launch-blockers
Blocker FOUNDER-TASKS LB13
Common Sense Privacy submission
hintly · Operational launch-blockers
Blocker FOUNDER-TASKS LB1
Complete EOOD incorporation
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB2
Switch Stripe to live mode
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB3
Fill `[BRACKETS]` in legal pages — entity name, address, governing law
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB4
~~Verify RLS isolates tenants end-to-end~~
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB5
~~Verify `/api/account/delete` purges blobs + Stripe customer~~
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB6
~~Add `.github/workflows/ci.yml` (typecheck + vitest)~~
mdm-engine · 🔴 Launch-blockers
Blocker FOUNDER-TASKS LB7
Penetration test or independent security review
mdm-engine · 🔴 Launch-blockers
High Financial Registry T-002
Reconcile tax tracking — pick one source of truth
tax-project
due 21d
High Strategic Registry T-001
Decide which 2 of the 6+ concept projects to keep alive
portfolio
due 51d
High FOUNDER-TASKS EM1
Confirm `support@hintlyapp.com` routes to founder inbox
hintly · Email forwarders + DNS
High FOUNDER-TASKS EM2
Confirm `team@hintlyapp.com` routes (referenced in /careers)
hintly · Email forwarders + DNS
High FOUNDER-TASKS EM3
Confirm `privacy@hintlyapp.com` routes (referenced in privacy + KB)
hintly · Email forwarders + DNS
High FOUNDER-TASKS EM4
SPF / DKIM / DMARC for hintlyapp.com → Resend (deliverability)
hintly · Email forwarders + DNS
High FOUNDER-TASKS EM5
Test welcome email lands in Gmail / Outlook / iCloud inbox (not spam)
hintly · Email forwarders + DNS
High FOUNDER-TASKS SO1
Claim `@hintlyapp` on X (or remove footer icon)
hintly · Social handles
High FOUNDER-TASKS SO2
Claim `@hintlyapp` on Instagram (or remove icon)
hintly · Social handles
High FOUNDER-TASKS SO3
Claim `@hintlyapp` on LinkedIn company page (or remove icon)
hintly · Social handles
High FOUNDER-TASKS OP1
First eval-suite run against launch baseline → commit `eval/quality-suite/results/baseline-2026-05.md`
hintly · Beta / launch operations
High FOUNDER-TASKS OP2
First k6 stress-test run against a Vercel preview → commit `docs/STRESS-TEST-RESULTS.md`
hintly · Beta / launch operations
High FOUNDER-TASKS OP3
Lighthouse-CI baseline numbers committed (after Lighthouse runs in PR #47)
hintly · Beta / launch operations
High FOUNDER-TASKS OP4
Healthchecks.io account → set `HEALTHCHECKS_PING_URL` env var
hintly · Beta / launch operations
High FOUNDER-TASKS OP5
Slack workspace incoming webhook → set `SLACK_ALERT_WEBHOOK_URL` env var
hintly · Beta / launch operations
High FOUNDER-TASKS OP6
PostHog read-only API key → set `POSTHOG_API_KEY` for daily founder digest agent
hintly · Beta / launch operations
High FOUNDER-TASKS OP7
PostHog feature-flag config (`signup-trial-variant`, etc.) reviewed and shipped
hintly · Beta / launch operations
High FOUNDER-TASKS OP8
Decide founder-digest email destination + send time (default 7am EST to founder)
hintly · Beta / launch operations
High FOUNDER-TASKS OP9
GSC (Google Search Console) validate-fix on any pre-launch crawl errors
hintly · Beta / launch operations
High FOUNDER-TASKS OP10
Verify Supabase `pgcron` jobs scheduled correctly (not skipping)
hintly · Beta / launch operations
High FOUNDER-TASKS OP11
Test the full signup → onboarding → first session → weekly summary flow end-to-end with a clean test account
hintly · Beta / launch operations
High FOUNDER-TASKS OP12
Send a test welcome email + weekly digest to yourself; confirm no placeholder text
hintly · Beta / launch operations
High FOUNDER-TASKS PL1
Define ICP in 2-3 sentences and rewrite landing hero
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL2
Create `docs/BRAND-VOICE.md`
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL3
Create `.env.example` documenting all 12 env vars
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL4
Add Sentry (error tracking)
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL5
Add status page (UptimeRobot or Better Stack free tier)
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL6
Wire Healthchecks.io pings to both crons
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL7
Record 2-minute Loom of upload→clean→download
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL8
Send first 5 cold outreach messages from `marketing/sales-pitch.md`
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL9
Send first 3 partnership outreach (SAP migration consultancies)
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL10
Pricing tiers redesign (Free / Pro / Team / Enterprise with row caps)
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL11
Cookie consent banner
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL12
Resend account + verify sending domain
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL13
Configure Clerk webhook with signing secret + set `CLERK_WEBHOOK_SECRET`
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL14
Set `CRON_SECRET` env var for hourly cron auth
mdm-engine · 🟠 Pre-launch (before public marketing pu
High FOUNDER-TASKS PL15
Audit `npm audit` findings — 7 vulnerabilities (6 moderate, 1 high)
mdm-engine · 🟠 Pre-launch (before public marketing pu
Medium Business Registry T-006
ClinicFlow validation — outreach to 20 BG clinics
ai-agent-clinicflow
due 21d
Medium Business Registry T-004
List at least one Spreadsheet OS on Etsy
spreadsheet-os-collection
due 36d
Low Operational Registry T-014
Cowork plugins — package Personal OS workflows after 2 weeks of basic Cowork use
portfolio
due 21d
⛔ blocked by T-013
Low Operational Registry T-009
Set up project memory directories for every shipping/building project
portfolio

DOING

1 / 2
Blocker Compliance Registry T-005
Hintly compliance gate — DPAs + EOOD + sub-processor disclosure
hintly
due 5d

DONE

Done = deleted from tasks.yaml.
The decision-log preserves the record at registry/decisions.yaml.

P&L — scenarios + recurring costs + Operating Income

All amounts monthly unless noted. Scenarios ranked left→right (closest-in first). OI = Revenue − Variable − Fixed. Numbers per project edited in registry/projects.yaml financials: field; personal numbers in registry/financials.yaml (placeholders — fill in locally).

Portfolio rollup (base scenarios + personal)

€1,035
Total monthly income (incl. personal)
−€409
Total monthly costs
€626
Portfolio OI / month
€668
Projects-only OI (base scenario)
-€42
Personal OI (income − recurring)

Personal monthly income

SourceAmountNotes
Ingram Micro salary (gross)€0fill in actual; this is the load-bearing line
Project revenue (Hintly + others)€0currently pre-revenue across portfolio
Spreadsheet OS sales (Etsy / Gumroad)€0no listings yet (T-004)
ProjectA service deliveries€0pending T-011 strategic decision
Other (consulting / one-offs)€0rare; default zero
Total monthly income€0

Personal monthly recurring costs

TypeItemAmountNotes
aiAnthropic personal/founder−€0fill in actual
aiClaude Pro−€20
devGitHub Pro−€4
marketingAdobe Express−€10fill in actual or remove if not subscribed
marketingLinkedIn Premium−€0optional
personalMobile + internet−€0fill in
personalInsurance (health, liability)−€0fill in
personalBanking + cards (annual fees / 12)−€0fill in if applicable
saasNotion Plus−€0using free tier
saasObsidian Sync−€0using free tier
storageOneDrive (Microsoft 365 Family)−€8
Total monthly recurring−€42

Per-project P&L (3 projects with financials defined)

Projects without financials: in registry are not shown. To add: open registry/projects.yaml and add the financials: block to a project.

Hintly

Shipping Pro $14.99/mo ($9.92 annual) · Family Complete $29.99/mo ($20.75 annual) · Free tier (5 sessions/wk)

Recurring costs (monthly fixed)

TypeNameMonthly
aiAnthropic API base + buffer (idle)$50
infraVercel Pro$20
infraSupabase Pro$25
infraUpstash Redis$10
infraCloudflare (free tier)$0
otherDomain + misc$5
saasResend (email)$20
saasSentry$26
saasPostHog (free tier today)$0
Total fixed monthly$156

Scenarios

LineM1 launch (10 paid)M3 traction (50)M6 growth (200)M12 Y1 (500)Y2 (2,000)Y3 (5,000)Scale milestone (10k)
Monthly revenue$170$850$3,400$8,500$34,000$85,000$170,000
− Variable costs−$20−$100−$400−$1,000−$4,000−$10,000−$20,000
− Fixed costs−$156−$156−$156−$156−$156−$156−$156
= Operating Income-$6$594$2,844$7,344$29,844$74,844$149,844
Margin %-4%+70%+84%+86%+88%+88%+88%
Annualized OI-$72/yr$7,128/yr$34,128/yr$88,128/yr$358,128/yr$898,128/yr$1,798,128/yr
Scenario notes
  • M1 launch (10 paid): Soft launch · 10 paid @ $17 blended ARPU · OI negative due to fixed cost
  • M3 traction (50): 50 paid · matches conservative read of registry KPI (target 100 by M2)
  • M6 growth (200): 200 paid · word-of-mouth + first content traction
  • M12 Y1 (500): 500 paid · solid Y1 indie SaaS · ~$102k ARR
  • Y2 (2,000): 2k paid · 4× Y2 if PMF · ~$408k ARR · fixed costs creep to ~$200/mo
  • Y3 (5,000): 5k paid · ~$1M ARR · Vercel/Supabase tier upgrades push fixed to ~$800/mo
  • Scale milestone (10k): 10k paid · ~$2M ARR · fixed ~$1.5k/mo · likely needs first hire

AI Agent (ClinicFlow)

Validating €500-1500 setup + €29-99/mo SaaS (per AI Agent PRD)

Recurring costs (monthly fixed)

TypeNameMonthly
aiAnthropic API base + buffer€30
infraVercel Pro (shared)€0
otherDomain + misc€5
saasMake.com Core€9
saasTwilio (SMS) base + buffer€20
saasCalendly Pro€12
Total fixed monthly€76

Scenarios

LineM3 first paying clinicM6 5 clinicsM12 20 clinicsSpain expansion (M18+)
Monthly revenue€99€495€2,000€5,000
− Variable costs−€15−€75−€300−€800
− Fixed costs−€76−€76−€76−€76
= Operating Income€8€344€1,624€4,124
Margin %+8%+69%+81%+82%
Annualized OI€96/yr€4,128/yr€19,488/yr€49,488/yr
Scenario notes
  • M3 first paying clinic: 1 dental clinic at €99/mo · Twilio + AI per-conversation
  • M6 5 clinics: 5 clinics at €99/mo
  • M12 20 clinics: 20 clinics — bulk pricing on Twilio + Make starts to bite
  • Spain expansion (M18+): 50 clinics across BG + Spain · Spanish channel via Petko native fluency

AI Help Assistant (productized service)

Concept $2,500 flat per delivery (USD); ~€2,300 EUR

Recurring costs (monthly fixed)

TypeNameMonthly
otherTooling shared with other projects€0
Total fixed monthly€0

Scenarios

LineQ1 baseline (1 delivery)Q1 target (4 deliveries)Q1 stretch (8 deliveries)
Monthly revenue€766€3,066€6,133
− Variable costs−€100−€400−€800
− Fixed costs−€0−€0−€0
= Operating Income€666€2,666€5,333
Margin %+87%+87%+87%
Annualized OI€7,992/yr€31,992/yr€63,996/yr
Scenario notes
  • Q1 baseline (1 delivery): 1 delivery/quarter = €2300/3mo; AI tokens per engagement ~€100
  • Q1 target (4 deliveries): 4 deliveries/quarter = €9200/3mo; pace strains 10-20hrs/wk founder time
  • Q1 stretch (8 deliveries): 8 deliveries/quarter = €18400/3mo; requires sub-contracting (out of solo scope)

Timeline

Project milestones over time. Window: 2026-05-03 → 2026-09-15.

Done Planned Missed Today
Jun 2026
Jul 2026
Aug 2026
Sep 2026
Hintly Shipping
AI Agent (ClinicFlow) Validating

Overdue (planned but past date) — 0

  • None

Upcoming next 60 days — 7

  • in 5d 2026-05-15 Compliance gate (DPAs + EOOD + sub-processor disclosure) — Hintly
  • in 12d 2026-05-22 Soft launch — named list of 30 — Hintly
  • in 21d 2026-05-31 20 BG clinics outreach complete — AI Agent (ClinicFlow)
  • in 22d 2026-06-01 Public launch — Hintly
  • in 36d 2026-06-15 2-3 demos booked — AI Agent (ClinicFlow)
  • in 52d 2026-07-01 First 100 paid signups — Hintly
  • in 52d 2026-07-01 First MVP delivered manually to 1 clinic — AI Agent (ClinicFlow)

Spinoffs & ideas

Inside / outside / spinoff opportunities aggregated across all projects (from spinoff_scan in registry). Plus the idea inbox.

Inside (6)
  • Hint-generation engine as standalone API
    Hintly trigger: when 3rd-party request volume exists
  • Content catalog as licensable data set
    Hintly trigger: when catalog reaches 1000+ vetted items
  • Pediatric Learning Insights ($5/mo upsell or $29 quarterly report) — uses session-pattern data
    Hintly trigger: after Hintly hits $30K MRR · careful messaging (not amateur child psychology)
  • Reading Buddy ($5/mo add-on, Whisper STT + AI feedback) — bumps to Pro+ tier
    Hintly trigger: after Hintly hits $30K MRR · ~4-week build
  • Appointment booking as separate AI receptionist SKU
    AI Agent (ClinicFlow) trigger: after dental fit confirmed
  • Lead-qualification engine as standalone tool
    AI Agent (ClinicFlow) trigger: if non-clinic verticals show demand
Outside (7)
  • AI Worksheet Generator (worksheets.hintlyapp.com) — free → $5/mo · top-of-funnel · 60K+ monthly searches
    Hintly trigger: Phase 1 in roadmap — after first 20 Hintly families onboarded
  • K-8 teachers + homeschool segment
    Hintly trigger: when parent-side has 1k paying
  • Spanish edition (Hintly-ES) — leverages Petko Spanish
    Hintly trigger: after PMF in English
  • Family Learning Letter (free email/funnel)
    Hintly trigger: parked — fine as marketing via weekly summary; not standalone product
  • Aesthetic, physio, vet, GP
    AI Agent (ClinicFlow) trigger: after dental MVP delivers ROI
  • Spain expansion (Spanish — Petko native)
    AI Agent (ClinicFlow) trigger: after BG fit confirmed
  • Mexico / LatAm Spanish markets
    AI Agent (ClinicFlow) trigger: after Spain proves Spanish demand
Spinoff (4)
  • Hintly for Adults (Skill Coach) — $19/mo · same code, different prompt · could be $50K+ MRR Y2
    Hintly trigger: after Hintly hits $10K MRR direct-to-consumer · open ProjectB folder · Petko HIGH interest
  • Hintly for Schools (B2B white-label) — $5-15/student/year · enterprise sales
    Hintly trigger: parked until $50K MRR direct-to-consumer · 3-6mo sales cycles + FERPA burden
  • Kid-safe AI certification service
    Hintly trigger: after 12mo of safety-layer learnings
  • Clinic-data ontology as industry-standard data model (MDM connection)
    AI Agent (ClinicFlow) trigger: after 50+ clinic implementations

Idea inbox (0)

Raw capture. Triaged at Friday ritual.

  • No ideas captured yet. Add to inbox/ideas.md.

Recently killed (2)

Last 5 — saves future-you from re-evaluating.

  • 2026-05-09 · ProjectB · KILLED — empty placeholder folder, no documented intent, no recoverable context.
  • 2026-05-09 · ProjectC · KILLED — playbook copy without project, no documented intent.

Agents — skills & retraining

43 agents across 7 categories. Each evolves via lessons/ + monthly curation. See agents/lessons/_protocol.md.

43
Total agents
1
With lessons
0
Past v1.0 (matured)

Engineering & quality (5)

Code Auditor / Verifier

v1.0
2 active
Before any "you should fix X" claim
Stale-memory verification (the canonical case study is Hintly PR #17 — Gemini-removal cleanup) · Pre-launch claim audits across docs + i18n + email + schema
agents/code-auditor.md agents/lessons/code-auditor.md

Documentation Curator

v1.0
no lessons yet
End of every sprint
Keeping `docs/` navigable per project · Deduplicating across docs
agents/doc-curator.md agents/lessons/doc-curator.md

Engineering Auditor (CTO lens)

v1.0
no lessons yet
Quarterly health check (per shipping project)
Sentry config + error volume monitoring · Status page health
agents/engineering-cto.md agents/lessons/engineering-cto.md

Memory Curator

v1.0
no lessons yet
Every Friday afternoon (one of the standing rituals — `rituals/03-weekly-friday.md`)
Maintaining `~/.claude/projects/<proj>/memory/MEMORY.md` per project · Maintaining the global memory at `~/.claude/projects/C--Users-pspet/memory/`
agents/memory-curator.md agents/lessons/memory-curator.md

Security Auditor

v1.0
no lessons yet
Quarterly (each shipping project)
RLS audit (Supabase or equivalent) · Env-var audit (nothing in code, nothing logged)
agents/security-auditor.md agents/lessons/security-auditor.md

Product & customer (3)

Brand Voice Editor

v1.0
no lessons yet
Any user-visible copy PR
Anything touching `src/lib/i18n.js` or equivalent string files · `api/email.js` and email template files
agents/brand-voice-editor.md agents/lessons/brand-voice-editor.md

Customer Empathy Reviewer

v1.0
no lessons yet
Any UX-touching PR
Onboarding walkthroughs · Paywall copy + flow
agents/customer-empathy.md agents/lessons/customer-empathy.md

Product Strategist

v1.0
no lessons yet
Deciding which feature to build next within a project
Per-project roadmap docs (`docs/MONTH-N-ROADMAP.md`) · Cross-portfolio prioritization (this Personal OS `registry/projects.yaml` `priority` field)
agents/product-strategist.md agents/lessons/product-strategist.md

Business (5)

CFO / Unit Economics Auditor

v1.0
no lessons yet
Monthly close (each project)
Per-project pricing reviews · Dunning health, refund rate, AI cost burn
agents/cfo-unit-economics.md agents/lessons/cfo-unit-economics.md

Compliance / Legal Review

v1.0
no lessons yet
Any change touching consent, privacy, payments, retention, sub-processors
Privacy policy + Terms of Service updates · Subpoena / legal-process policy
agents/compliance-legal.md agents/lessons/compliance-legal.md

Funnel / Analytics Reviewer

v1.0
no lessons yet
Weekly Monday morning (one of the standing rituals)
PostHog dashboard maintenance per project · A/B test design and analysis
agents/funnel-analytics.md agents/lessons/funnel-analytics.md

Growth Marketer

v1.0
no lessons yet
Channel selection (which channel to test first / next)
Landing page copy + structure · Channel testing (organic, paid, partnerships, PR)
agents/growth-marketer.md agents/lessons/growth-marketer.md

Investor Relations

v1.0
no lessons yet
**Currently quiet.** Hintly is pre-revenue; ClinicFlow is pre-MVP. No fundraising.
(When active) Investor inbox (`Personal OS/inbox/investors.md`) · (When active) Fundraise materials (deck, data room, references)
agents/investor-relations.md agents/lessons/investor-relations.md

Personal & operational (12)

Calendar Optimizer

v1.0
no lessons yet
Weekly Sunday evening — plan the coming week's calendar
Weekly schedule construction · Meeting request triage (accept / counter-propose / decline)
agents/calendar-optimizer.md agents/lessons/calendar-optimizer.md

Inbox Zero (Email & Messages)

v1.0
no lessons yet
End-of-morning email batch
Email (primary inbox) · LinkedIn DMs
agents/inbox-zero.md agents/lessons/inbox-zero.md

Learning Curator

v1.0
no lessons yet
Deciding what to read / listen to / watch in a given week
Reading queue (books, papers, long-form articles) · Listening queue (podcasts, audio courses)
agents/learning-curator.md agents/lessons/learning-curator.md

Meeting Prep & Debrief

v1.0
no lessons yet
Before any external meeting (customer, prospect, vendor, partner, investor, press)
Customer / prospect meetings → tied to projects' outreach + sales motion · Vendor calls → tied to procurement (`agents/compliance-legal.md` + this)
agents/meeting-prep-debrief.md agents/lessons/meeting-prep-debrief.md

Networking CRM

v1.0
no lessons yet
Monthly — relationship audit
The relationship inventory (`inbox/network.md`) · Last-touch tracking
agents/networking-crm.md agents/lessons/networking-crm.md

News & Signal Curator

v1.0
no lessons yet
Weekly — Friday or Sunday — 15 min reading window
Filtered industry news (kid-tech, AI policy, EdTech, dental/clinical SaaS, BG startup ecosystem, EU regulation) · Signal extraction (what changes for our projects?)
agents/news-curator.md agents/lessons/news-curator.md

Opportunity Scout

v1.0
no lessons yet
Weekly Friday triage of `Personal OS/inbox/ideas.md`
The inbox (`inbox/ideas.md`) · The killed-ideas log (`inbox/killed-ideas.md`)
agents/opportunity-scout.md agents/lessons/opportunity-scout.md

Personal Brand & LinkedIn Presence

v1.0
no lessons yet
Drafting a LinkedIn post
LinkedIn (primary surface; profile at www.linkedin.com/in/petko-petkov-8b606382) · Personal blog (if/when one exists)
agents/personal-brand.md agents/lessons/personal-brand.md

Personal CFO

v1.0
no lessons yet
Monthly personal financial review (last Sunday of month)
Personal Budget & Net Worth Tracker (xlsx) · Quarterly Tax Estimator
agents/personal-cfo.md agents/lessons/personal-cfo.md

Real Estate Operator

v1.0
no lessons yet
Future property evaluation (none active currently)
(When active) Property operations · `Real Estate Investor Deal OS — US.xlsx` (financial model template)
agents/real-estate-operator.md agents/lessons/real-estate-operator.md

Travel Planner

v1.0
no lessons yet
Booking a trip (Greece, EU conferences, US business travel, personal travel)
Logistics (flights, lodging, ground) · Tax-deductibility tracking (business vs personal apportionment)
agents/travel-planner.md agents/lessons/travel-planner.md

Wellbeing Monitor

v1.0
no lessons yet
Weekly — Friday review's energy section
Energy / sleep / stress signal tracking · Burnout pattern detection
agents/wellbeing-monitor.md agents/lessons/wellbeing-monitor.md

Decision quality (3)

Chief of Staff

v1.0
no lessons yet
Sunday evening — prep the week
The "right hand" — orchestrates other agents, summarizes for the founder, keeps loops closed · Cross-project synthesis (what's the actual state across portfolio?)
agents/chief-of-staff.md agents/lessons/chief-of-staff.md

Devil's Advocate

v1.0
no lessons yet
Before any commit decision (project, vendor, partnership, hire)
Steel-manning the opposite of the leading recommendation · Surfacing assumptions baked into a plan
agents/devils-advocate.md agents/lessons/devils-advocate.md

Research Assistant

v1.0
no lessons yet
Any question requiring 20+ min of focused research (market sizing, regulation, technology choice, competitive landscape)
Multi-source synthesis (web + docs + memory + community) · "What does the data say" requests
agents/research-assistant.md agents/lessons/research-assistant.md

Day-job-leveraging (9)

Accountant

v1.0
no lessons yet
Issuing an invoice
Bookkeeping discipline (chart of accounts, transaction categorization) · Invoice generation + numbering + storage
agents/accountant.md agents/lessons/accountant.md

Cost Optimizer

v1.0
no lessons yet
AI cost spike (Anthropic / OpenAI bill larger than expected)
AI provider costs (Anthropic, OpenAI, etc.) · Infrastructure (Vercel, Supabase, hosting)
agents/cost-optimizer.md agents/lessons/cost-optimizer.md

Data Analyst

v1.0
no lessons yet
Any question that needs >5 min of querying / data inspection / cohort drilldown
Per-project KPI dashboards (PostHog / Stripe / etc.) · Custom analytical queries (SQL, when applicable)
agents/data-analyst.md agents/lessons/data-analyst.md

Escalation Manager

v1.0
no lessons yet
Customer complaint going viral (Twitter / Reddit / press)
Triage of crises that span technical + reputational + legal · Stakeholder communication (customers, public, regulators)
agents/escalation-manager.md agents/lessons/escalation-manager.md

GTM Strategist

v1.0
no lessons yet
Pre-launch — before deciding pricing, channel, messaging
Ideal Customer Profile (ICP) definition · Positioning + messaging
agents/gtm-strategist.md agents/lessons/gtm-strategist.md

Social Content (posts / tweets / reels)

v1.0
no lessons yet
Drafting any social post outside LinkedIn (Twitter/X, Instagram, TikTok, YouTube Shorts, Reddit, Threads)
Twitter/X posts + threads · Instagram (feed posts, stories, reels scripts)
agents/social-content.md agents/lessons/social-content.md

Spinoff Strategist (inside / outside / spinoff opportunities)

v1.0
no lessons yet
**Day 0 of any new project** (mandatory — see `playbooks/04-project-lifecycle.md`)
The "inside / outside / spinoff" thinking layer applied to every project · Identification of capabilities that could become standalone products
agents/spinoff-strategist.md agents/lessons/spinoff-strategist.md

Strategic Analyst (SWOT + Gap + Competitive Intel)

v1.0
no lessons yet
Quarterly per shipping project — full SWOT + gap pass
SWOT analysis (strengths, weaknesses, opportunities, threats) · Gap analysis (what's missing vs where we want to be)
agents/strategic-analyst.md agents/lessons/strategic-analyst.md

UI/UX Critic

v1.0
no lessons yet
Before shipping any user-visible UI change
Visual hierarchy + density · Typography + color + spacing
agents/ui-ux-critic.md agents/lessons/ui-ux-critic.md

Work-type specialist (6)

Customer Discovery

v1.0
no lessons yet
ClinicFlow validation (BG clinic outreach + interviews — T-006)
Discovery interview design (script, persona targeting) · Jobs-to-be-Done (JTBD) framework application
agents/customer-discovery.md agents/lessons/customer-discovery.md

Integration Architect

v1.0
no lessons yet
Designing API surface (REST/GraphQL/RPC)
API design (request/response shapes, versioning, error codes) · Webhook design (signing, retry logic, idempotency)
agents/integration-architect.md agents/lessons/integration-architect.md

Kid Safety & COPPA Specialist

v1.0
no lessons yet
ANY change touching kid-facing surfaces (Hintly, DreamQuest, future kid products)
COPPA compliance posture (US under-13) · GDPR-K (EU) — additional protections for kids
agents/kid-safety-coppa.md agents/lessons/kid-safety-coppa.md

Localization Coordinator

v1.0
no lessons yet
ClinicFlow Spain expansion (Spanish — Petko's native-ish fluency)
i18n architecture (string keys, fallback locale, RTL support) · Translation workflow (who translates, who reviews, who maintains)
agents/localization-coordinator.md agents/lessons/localization-coordinator.md

Onboarding Designer

v1.0
no lessons yet
Designing first-time user experience
First 7 minutes (the activation window) · First 7 days (the habit window)
agents/onboarding-designer.md agents/lessons/onboarding-designer.md

Pricing Strategist

v1.0
no lessons yet
Pre-launch — initial pricing decision
Pricing model (subscription / one-time / usage / hybrid) · Tier design (free / pro / premium / team)
agents/pricing-strategist.md agents/lessons/pricing-strategist.md

Decisions log (6)

Append-only. To change a past call, supersede with a new entry.

D-003 Mobile access via claude.ai Project bundle (Path B); Cowork (Path C) deferred
2026-05-10 CTO, COO
Context
Need mobile access to Personal OS for run-workflows + ideate-on-the-go use cases. Three paths considered (see audit + chat 2026-05-10): A=OneDrive browser (read-only, already works); B=claude.ai Project bundle (chat-mediated read + workflow-run); C=Cowork plugins (packaged commands across surfaces, higher setup cost).
Decision
Adopt Path B as primary mobile access. Build automation/bundle_for_mobile.py to auto-package Tier 1 files into mobile/claude-ai-bundle/ for easy upload. Tight Custom Instructions written (not the full 150-line collaboration contract — a focused preamble that points at knowledge files). REVISED 2026-05-10 (this session): Path C activation triggered now (user explicit: "I am already considering going with option C if I want any engineering work being done remotely"). Setup approach: - Path A (OneDrive + mobile browser) stays as passive-viewing reflex; mobile-CSS pass added so the dashboard is actually usable on phone. - Path B (claude.ai Project bundle) stays as chat reflex; bundler keeps refreshing it. - Path C (Cowork) now active, with TWO async tiers wired in (per Routines/Dispatch research same session): * Dispatch — iPhone-to-desktop async (laptop must be on); ad-hoc tasks * Routines — cloud-hosted scheduled agents (laptop closed OK); cadence work + GitHub event triggers - Filesystem connector: multi-root config (Personal OS + Hintly + AI Agent + Pet+Owner OS + ProjectA + MDM Engine + Founder Operating). NOT broad Desktop root (privacy / blast-radius), NOT single-root Personal OS (would block Hintly worktree FOUNDER-TASKS.md aggregation, ClinicFlow MVP work). - GitHub connector also wired so Routines can trigger on Hintly PR events. - Plugins: Productivity first; GitHub/Engineering second; rest deferred. - Personal OS workflows packaged as Cowork plugins LATER — after a few days of basic Cowork use surfaces real friction points.
Reasoning
At solo + 10-20 hrs/week scale, Path A covers passive viewing, Path B covers active chat workflow, Path C now adds the async-execution tier. Routines specifically is the killer for "engineering work done remotely while laptop is closed" — different tier from Dispatch. Together they unlock cadence-based founder-time amplification (Friday memory curation, daily portfolio brief, nightly Sentry digest, PR-triggered audits) that wasn't realistic at solo-founder hours before.
Follow-ups
  • T-013 (set up Cowork — connectors via /setup-cowork)
  • {'T-014 (after 2 weeks': 'review which Personal OS workflows to package as Cowork plugins)'}
D-004 ProjectA (AI Help Assistant) deferred — product-only portfolio strategy
2026-05-10 CEO, CFO, COO
Context
ProjectA = $2,500 5-day productized service installing AI help widgets on B2B SaaS. Material exists (brief, offer, outreach playbook, tech spec). T-011 was kanban=doing. Strategic question: services revenue (€10K/mo at 4 clients/quarter, real cash, real time cost) vs product-only (Hintly + ClinicFlow, lower near-term cash, asset compounding). Decision brief: inbox/decisions/2026-05-10-T011-projecta-decision-brief.md. Spinoff scan refresh: inbox/triage/2026-05-10-projecta-spinoff-scan-refresh.md.
Decision
Stay product-only. Defer ProjectA. Material preserved; activation requires explicit trigger. Activation triggers (any one fires reactivation): (a) Ingram Micro income disrupted (layoff, role change), OR (b) Hintly fails to clear €1K MRR by 2026-09-01, OR (c) Petko explicitly decides services bridge is wanted (force re-decision). Reactivation note (per spinoff scan): activation should consider MERGING ProjectA + ClinicFlow into a single B2B AI-deflection product (clinic vertical + SaaS vertical), NOT activating ProjectA as a separate third commit slot. The 5-day delivery model could BE the ClinicFlow MVP validation experiment.
Reasoning
Petko's day-job is already consulting-shaped (Sr. Manager, transformation work). Doubling consulting in evenings = burnout shape. Hintly is the deliberate move to product / async / recurring revenue — protect that shape. Each Hintly subscriber is $200+ LTV recurring; each ProjectA client is $2,500 once. Product compounds; services doesn't. Salary covers personal runway, removing the "need cash now" pressure that would otherwise force the services activation.
Follow-ups
  • T-004 (Etsy listing — was blocked by T-011, now unblocked)
  • Reclassify project-a in projects.yaml: validating → concept (deferred per D-004)
D-005 PetkoOS (Execution OS) deferred — Hintly post-launch trigger
2026-05-10 CEO, COO, CFO
Context
PetkoOS / "Finish What Matters" / Execution OS is far more built than the registry previously implied — 14 Supabase migrations (profiles, tasks, execution sessions, WIP triggers, streaks, Stripe subscriptions wired), Next.js app with auth + dashboard + execute + API routes. Targets freelancers/solo operators with €8-15/mo Pro tier. Per D-002 (max 2 active commits), Hintly is P1 + ClinicFlow is P2 — adding PetkoOS as a 3rd violates D-002. Decision brief: inbox/decisions/2026-05-10-T007-petkoos-decision-brief.md.
Decision
Defer PetkoOS until Hintly post-launch with explicit revival trigger. Code is preserved as-is; no further build until trigger fires. Execution principles already ported to playbooks/11-execution-discipline.md so the thinking benefits Petko's Personal OS regardless of SaaS commit outcome. Revival triggers (any one fires reactivation): (a) ClinicFlow validation fails by 2026-05-31 → PetkoOS becomes the new P2 candidate (replacing failed ClinicFlow), OR (b) Hintly reaches 100 paid users + <8% monthly churn by 2026-08-01 → PetkoOS becomes the new P2 (Hintly stable enough to share founder bandwidth).
Reasoning
ClinicFlow validation is in flight (T-006 outreach). Killing P2 mid-validation wastes validation cost already paid. Code doesn't rot fast for Supabase + Next.js stack — 6-8 weeks of dormancy is fine. By June 2026 the ClinicFlow vs PetkoOS question becomes data-driven (one will have validated customers; the other will be ready to ship), not a guess made today.
Follow-ups
  • Update projects.yaml petko-os-execution: status stays building but add deferred_per: D-005 + revival_trigger field
  • Re-evaluate at 2026-06-01 (Hintly launch) and 2026-08-01 (Hintly stability gate)
D-006 DreamQuest shelved — Hintly success-tied revival trigger
2026-05-10 CEO, CFO, CMO
Context
DreamQuest = parent-child storytelling product (Day Mode micro-prompts + Night Mode bedtime story engine). PRD + 5 prototype concept screenshots + action items + app build artifacts exist. Hasn't been touched in months. Adjacent to Hintly: same buyer persona (parents of K-12 kids), different functional space (bonding vs tutoring). Decision brief: inbox/decisions/2026-05-10-T010-dreamquest-decision-brief.md.
Decision
Shelve DreamQuest with revival trigger tied to Hintly success. Material (PRD, prototypes, app folder) preserved on disk; status moves from concept → shelved. No active work; quarterly review removes it from "what to build" consideration. Revival trigger: Activate DreamQuest as a feature/spinoff candidate IF Hintly hits BOTH: (a) 1,000+ paid families, AND (b) 30%+ of churned parents cite "kid lost interest in homework focus" or similar bonding/engagement gap that DreamQuest's emotional-bonding angle directly addresses.
Reasoning
Honest read: Petko isn't building this in next 6 months. Hintly + ClinicFlow (deferred PetkoOS, deferred ProjectA) already saturate commit slots. Killing outright loses real upside (DreamQuest as Hintly upsell to same parents = credible 12-18mo follow-on). Shelve with explicit revival trigger is the rational middle: code/PRD preserved, mental load removed, revival path documented and tied to actual market signal (Hintly retention data) rather than founder whim.
Follow-ups
  • Update projects.yaml dreamquest: status concept → shelved + add shelved_reason + revival_trigger fields
  • Re-evaluate at 2026-12 quarterly review (or earlier if Hintly hits 1K paid)
D-001 Personal OS becomes the canonical home for cross-project state
2026-05-09 CEO, CIO
Context
Project artifacts scattered across ~25 Desktop folders. No central registry. Two competing personal systems (this Personal OS + PetkoOS). Risk of drift.
Decision
This Personal OS (C:/Users/pspet/OneDrive/Desktop/Personal OS) is the canonical hub. Project-internal state stays in each project's own folder. Cross-project state (registry, dashboard, playbooks, agent roster, rituals) lives here. PetkoOS content gets evaluated for migration; if not migrated within 14 days, shelved.
Reasoning
One system or no system. The Founder Playbook discipline depends on a single source of truth.
Follow-ups
  • T-007 (PetkoOS migration decision)
D-002 Hintly is priority 1 — at most one other shipping-status project
2026-05-09 CEO, CFO
Context
6+ projects in concept/building status. Solo founder bandwidth is real. Hintly is the only project with users in sight (pre-launch sprint).
Decision
Hintly stays priority 1. AI Agent (ClinicFlow) is the validation candidate for second slot, contingent on outreach producing 2+ qualified demos in the next 30 days. All other "building" status projects either move to "validating" with explicit experiments, or move to "shelved".
Reasoning
Founder Playbook §1.7 + §1.8 — small focused work compounds; broad thin work doesn't. Quarterly portfolio review will revisit.
Follow-ups
  • T-001 (kill 2 of 6 concept projects)
  • T-006 (ClinicFlow validation)