June 25 for Fred. One GTM operating system, two lanes riding it (coach + Dragon 2). Every stage below is a live, clickable artifact, not a slide. Nothing is in production, nothing was sent without you.
π Jun 25 2026 Β· end of day
Three live surfaces. Click any one.If it were slop, these would not load with real data.
build command to live draft (not sourcing or copy)
3 live
coach campaigns IN_PROGRESS from the launcher
485274
NL relatietherapeuten, sending now
dragon2
protocol site live in preview
0
production deploys β all gated to you
clean
honesty + leak gates enforced in CI
#13 Β· #76
PRs submitted + tested
0 sends
Scalantec automation lane, ~1 month
The operating system Β· this spine is the live GTM map
One machine, seven stages, source to report.
The canonical memory bank generates the surface; the funnel turns every event into CRM state; the dashboard reports it; a feedback loop runs the coach page back into the CRM. The whole thing ships behind controlled-release rails with you holding the irreversible steps. The same seven stages are rendered live on the GTM map β this board is the narrated version of it.
Controlled release: leak-check β release, CI as the required checkNo direct prod accessFounder holds the irreversible gates
The two lanes Β· same machine, opposite stealth
Lane A
Coach
Consumer. Manykind-stealth: nothing served names Cere.
Shipped
A ~1-minute launcher (one build sweep, one gated start) plus an orchestrator so Filip drives the whole loop up to the two gates I hold.
Surface
Stealth-safe site, variants + personalized /c pages, leak-check on every release. CEF-AI/manykind-website Β· PR #13
Proof
Three campaigns IN_PROGRESS from the same skill (repeatable, not a one-off). Latest: 485274 to 5 real NL relatietherapeuten, honesty-gated Dutch sequence, written back to the CRM. IN_PROGRESS485274 Β· NL485104 Β· Christian485088 Β· niche test
Honest bound
Sent, not yet converted. No replies or calls claimed β conversion is gated by the demo clip (stage 06).
Your gate
Deploy + start. Nothing sends without you.
Lane B
Dragon 2
Protocol operators. Public Cere: the front door, named openly.
Shipped
Protocol home with the Dragon 1 story and a Join Dragon 2 path, plus operator intake into the same CRM (program=dragon-2-eu-gpu).
Redirect + both CTAs verified, zero Founding 50 leaks. Email lifecycle built but dark. nothing sends yet Canonical brief: CEF-AI/company-memory-bank Β· PR #76
Your gate
Reward copy β production cutover. Untouched by design.
Fred will ask Β· pre-answered, each links to proof
Isn't this just AI slop?
No, and it's one click to check. The map, the dashboard, and the CRM are live and load real data. It's one architecture reused across two lanes, not scattered output. Slop overclaims; every number here states its own bound.
A campaign in one minute, really?
The minute is the build command only (list + leads + draft + sequence), not sourcing or copywriting. The point isn't the minute, it's that launching stopped being the bottleneck. Sourcing, copy, and your gate are what remain.
Is it real, or a demo? Any replies?
485274 is verified IN_PROGRESS, real connection requests to 5 named NL therapists, in the CRM. I claim sent, not converted β no replies or calls asserted. Conversion is gated by the demo clip.
Why is Scalantec at 0 after a month?
Not competency. 482503 never started, EmailBison never ran, 472902 was wired to an empty list. Real blockers existed and the team pushed campaigns out by hand. The fact that stands: 0 production volume at β¬3,900/mo.
Does this scale past Martijn?
Yes. The /coach-campaign skill makes the loop Filip-runnable end to end; you hold only deploy + start. It's a repeatable system, evidenced by three live campaigns, not a hero run.
Will it embarrass us / overclaim?
The honesty gate is enforced in code (the build aborts on banned terms, "voice", earnings, em-dashes) and leak-check blocks any Cere mention from shipping. Brand integrity is a gate, not a hope.
The contrast, said precisely
Solo lanes shipped. The contractor lane did not.
Three live campaigns and two real surfaces went out today from the skill-based lanes. The contractor automation lane (β¬3,900/mo) is still at 0 sends: HeyReach 482503 never started, EmailBison never ran, 472902 was wired to an empty list. Real blockers existed and the team is capable, so this is an execution-and-blocker gap, not competency. The hard fact stands: zero production volume after a month.
What's on you
The few calls only you make
Coach: the demo clip is the conversion bottleneck (reply β call), and decide where the stealth tooling lives (private, not the shared installer).
Dragon 2: reward-copy decision gates the cutover, PR #76 merge is the hard gate, and the protocol site needs an org home.
Spend: keep funding the blocked automation lane, or fund the lane that shipped today.
Paste-ready message to Fred β Slack / email
Fred, the through-line on today. One operating system, two lanes.
Before you read it as a wall of AI output: it isn't, and it's one click to check. Open these three and watch them load real data.
- The map (how it wires together): https://manykind-gtm-map.pages.dev/
- The dashboard (the live funnel): https://manykind-coach-funnel.pages.dev/
- The CRM (the single source of truth): the Coach Contacts DB in Notion.
TL;DR: I shipped two go-to-market lanes today, the consumer coach lane and the Dragon 2 operator lane. They are not two projects. They are the same seven-stage machine pointed at two audiences. Both live in preview or test, nothing in production, nothing sent without you.
The machine, source to report: memory bank (canon) β generate (video + coach discovery) β build (site + personalized /c pages) β send (HeyReach launcher) β CRM (one source of truth) β convert (the demo-clip gate) β report (dashboard + #gtm digest). That exact spine is what the map renders live.
Lane 1, Coach (consumer, stealth): three campaigns are IN_PROGRESS from the same launcher, so this is repeatable, not a one-off. The latest, 485274, is live to 5 real NL relatietherapeuten with an honesty-gated Dutch sequence, written straight to the CRM. Honest bound: it is sent, not yet converted. I am not claiming replies or calls.
Lane 2, Dragon 2 (protocol operators, public): preview live at dragon2.scp-ecosystem.pages.dev, protocol home with the Dragon 1 story and the Join Dragon 2 path, operator intake into the same CRM. Email lifecycle built but dark. Production stays Founding 50 until the gated cutover.
Why it matters to you:
- One source of truth, fully measurable. Both lanes resolve to one CRM, so "is the agent converting" is answerable per campaign.
- QA and security on every release, no direct prod access, enforced in CI. The honesty gate is in code, not a promise.
- You hold the irreversible decisions. Coach: deploy + start. Dragon 2: cutover + first send.
- One agent, many faces. Hiring, marketing, sales run on the same machine. Coach outreach is the sales face of it.
- Stealth boundary intact. Coach surfaces never name Cere; the Dragon 2 surface is the public Cere front door. Same machine, opposite stealth.
The contrast worth seeing: the solo, skill-based lanes shipped three live campaigns and two surfaces today. The contractor automation lane is still at 0 sends after a month at EUR 3,900/mo.
What's on you:
- Coach: the demo clip is the conversion bottleneck (reply to call), and a call on where the stealth tooling lives.
- Dragon 2: the reward-copy decision gates the cutover, the PR #76 merge is the hard gate, the protocol site needs an org home.
- Spend: keep funding the blocked automation lane, or fund the lane that shipped today.