gtm.inc Launching soon
GTM — get to market, provably · beta

Know if your launch is ready. Then make it ready.

GTM scores your product across nine launch lenses, turns every weakness into an evidence-linked gap, and tasks your own coding agent to close them — then re-verifies and moves the score.

for founders shipping with a coding agent — Claude Code, Cursor, or any MCP agent · in private beta
Wrenfield — AI meeting-notes copilotREVIEW DONE
61/ 100
Beta gate · ≥654 POINTS SHORT
Launch gate · ≥75NOT YET
Web copy38CRITICAL Primary conversion path cannot be verified
Security72repo secrets scan unknown
User journeys76signup path unverified
Pricing80pricing transparency not assessed
unretouched: this one review wrote 42 findings and opened 8 fixes
The loop

Score → gaps → agent fixes → re-verify.

Not another dashboard you update by hand. The loop closes itself: every gap traces to evidence, every fix comes back through the same lenses that found it.

1
Capture
site · repo · docs
2
Review
lenses · evidence
3
Plan
gates · milestones
4
Execute
your agent · PRs
5
Re-measure
verify · re-score
readiness 7278 · re-baselined
Evidence, not vibes

Nine lenses. Real findings. A number you can defend.

SecurityW 13HIGHFRONTIERre-runs on code change
PricingW 13MEDIUMFRONTIER + WEBfortnightly
USP & positioningW 13MEDIUMFRONTIER + WEBfortnightly
Infra & releaseW 12MEDIUMSTANDARDre-runs on deploy
Journeys & first valueW 12HIGHFRONTIER + BROWSERre-runs on release
Feedback instrumentationW 10LOWSTANDARDmonthly
Web copy & languageW 9MEDIUMSTANDARD + WEBre-runs on site change
RisksW 9MEDIUMFRONTIERper review
Support pathwaysW 9LOWSTANDARDmonthly
From one live review
findings written42
generated by the LLM assessor12
gaps opened, ranked8
biggest single score impact−25
score after one fix cycle61 → 81
Absent data is reported absent — never faked as zero. If a lens can't verify something, it says so and scores accordingly.
38
Web copy & messaging — full lens report
Wrenfield · reviewed 2026-08-30 00:45 UTC · 19 findings, 12 shown · every row carries its evidence
headline31
conversion18
proof44
seo/meta62
CRITICALPrimary conversion path cannot be verified
hero CTA → /signup returns 404 · no capture form found on any linked page
conversion−25
HIGHHeadline names no outcome
"Meetings, organised." — no who, no result, no timeframe above the fold
headline−15
MEDIUMNo quantified value claim
zero numbers in hero or features · nearest metric is in footer smallprint
proof−8
MEDIUMSocial proof absent
no logos, quotes, counts or case links anywhere on the page
proof−8
MEDIUMDifferentiation unstated vs named category
page claims "unlike other note-takers" but never says what differs
headline−8
LOWMeta description missing
<meta name="description"> absent · SERP snippet will be arbitrary
seo/meta−4
LOWTitle tag generic
<title>Home</title> · carries neither brand nor category
seo/meta−4
LOWHero image has no alt text
3 of 7 images unlabelled · screen readers skip the product shot
seo/meta−2
LOWMobile hero truncates at 390px
headline clips mid-word on iPhone 14 viewport · CTA below fold
headline−2
INFOPricing page linked but empty
nav links /pricing · page returns placeholder copy · scored under Pricing lens
conversion0
INFOTone consistent across pages
voice analysis: 0.91 cosine similarity across 6 pages · no action needed
proof0
PASSHTTPS + canonical URLs correct
all links resolve https · rel=canonical present on every page
seo/meta
1 critical · 1 high · 3 medium · 4 low · 2 info · 1 passfix the top two → projected 38 → 57 · full lens ceiling 74
Risks & unknowns

Live risks — each with its next check.

Risks come from the lenses with evidence attached. Nothing sits on the register without a next check that retires or confirms it — sorted by exposure, expand for evidence and mitigation.

R1Admin route exposed pre-betaH×H · security lens
ExposureA data breach during the invite wave would end the launch — 25 hand-picked beta users are exactly the people who would tweet it.
Evidencecurl -s https://promptdeck.app/admin → HTTP 200, full admin panel, no auth header required.captured 14:18 Aug 22 · transcript on shelf
Mitigation on the planFix S-1 (briefed to your agent, ~2h) → security re-scan verifies → beta gate unblocks.
Next check — retires or confirmsSecurity re-scan after S-1 merges — scheduled as the blocking task before the beta gate. If clean, R1 retires.
Brief mitigationSee on planAccept risk…
R2Platform ships prompt versioning nativelyM×H · competitive shadow
ExposureThe platform absorbing your core feature flattens the differentiation story before wave 2.
Next check — retires or confirmscompetitor changelog watch · weekly · Reposition brief queued if shipped — USP lens re-run confirms.
Brief mitigationSee on planAccept risk…
R3$9 price traps the product lowH×M · pricing lens
ExposureAnchoring at hobby pricing makes the team tier unsellable later.
Next check — retires or confirms5 concierge offers out — will teams pay? · Pricing lens re-run after probe replies · retires on 2+ team yeses.
Brief mitigationSee on planAccept risk…
R4TTFV 11m 40s loses invite-wave usersM×M · journeys lens
ExposureHand-picked invitees who bounce once rarely return.
Next check — retires or confirmsjourney re-walk after onboarding fix J-1 · Retires under 4 minutes measured.
Brief mitigationSee on planAccept risk…
R5Solo-founder load in launch fortnightM×M · capacity
ExposureEverything on the plan lands through one pair of hands.
Next check — retires or confirmsagent-task throughput this week · Confirms if briefed tasks slip 3+ days.
Brief mitigationSee on planAccept risk…
R6Hobby-tier limits under launch spikeL×M · infra lens
ExposureFree-tier infra caps could brown-out the launch-day traffic.
Next check — retires or confirmsload probe before invite wave · Retires on clean 10× headroom.
Brief mitigationSee on planAccept risk…
R7Copy claims outpace the productL×M · web copy lens
ExposureTwo feature claims reference roadmap items, not shipped ones.
Next check — retires or confirmsclaims audit on next site change · Retires when claims match shipped surface.
Brief mitigationSee on planAccept risk…
Living plan · Gantt

Every gap, on a timeline to launch.

Reviews seed the plan automatically. Tasks carry their lens, severity and score impact; gates sit on the timeline as hard lines. Drag to re-baseline and GTM versions the snapshot — nothing is lost.

  • Beta & launch gates as timeline markers
  • Critical path & dependencies
  • Tasks a paired agent is actively working, flagged live
gtm.inc/plan
Launch planbaseline v3 · 6 weeks
TaskW1W2W3W4W5W6
beta 65
launch 75
Rewrite hero copy copy
done
Add /pricing page pricing
in progress
Rate limit public API security
agent
Add auth tests security
planned
Journey audit journeys
planned
Launch checklist gates
planned
Drag to re-baseline — versioned beta
Do the work

Task your agent. Or just fix the top thing.

gtm.inc/agents
Agent tasksclaude-code · auto-low-risk
Rate limit public API
brief #482 · security · +11 if merged
PR #482 open
Add /pricing page
brief #479 · pricing · +9 if merged
running
Rewrite hero for launch outcome
brief #471 · web copy · +4 if merged
merged
Add auth tests
brief #486 · security · ask-per-task
queued
gtm.inc/prioritise
Focus — do these nextranked by score gain
1
Rate limit public API
security · high · ~1h effort
+11Brief agent
2
Add /pricing page
pricing · high · ~2h effort
+9Brief agent
3
Rewrite hero for outcome
web copy · medium · ~30m effort
+4Brief agent
4
Run the journeys lens
journeys · never run · unknown
+?Run lens
The launch coach

Stuck on a gap? Ask the coach.

Grounded in your product graph, scores, plan and evidence — not generic advice. A shared thread your whole team can read, join and act from.

JJAL Beta timing — invite this week? SHARED THREAD
jj
Should I do the beta invite this week or wait?
JJ
CO
Coachgrounded
Wait — but days, not weeks. Readiness is 54; the beta gate needs 65. The three briefed fixes are worth ~+14 combined and all land this week if you approve the S-1 brief today. Inviting now sends 25 hand-picked people into a flow with an 11m 40s time-to-first-value — you'd burn your best contacts on a known-broken funnel. Target: approve S-1 now, invite Monday.
score 54 →gap S-1TTFV 11m40s
AL
alex
Agree with waiting — can we at least send the "you're on the list" warm-up email this week?
CO
Coachgrounded
Yes — that's pure upside and needs nothing from the fix queue. I've drafted it from the USP lens (outcome-first subject, no product claims you can't yet back). It's in the asset pack as "waitlist warm-up."
asset pack
jj
What would you cut if I only had 10 hours this week?
JJ
CO
Coachgrounded
Keep: S-1 (2h, +9), copy PR (1h, +6), pricing restructure (3h, +7). Cut: J-2 resend — wave 1 can live without it. That's 6 hours for +22 and the gate clears.
plan · 3 keptprojected 54 → 68
Ask, or comment for the team…Send
What the coach can see
product graph · all lens reports
score history · the plan + gates
feedback items · competitor set
revenue scenarios
It never sees another customer's data.
Suggested prompts
Draft my beta invite from the USP lensWhich gap closes fastest?Compare my pricing to the competitor setWhat did the last review change?
Revenue scenarios

Three futures, and the levers that move between them.

Scenarios are built by the pricing lens from your funnel; from first income, actuals plot inside the bands. Pull a lever to model a what-if and save it as a sub-scenario you can track against.

What-if modeller · pull a lever · save a set as a sub-scenario
Price /moBuilder tier$29
Deep-run COGScost to serve$12
New users /motop of funnel400
Free → paidconversion4%
Monthly churn12%
Freemiumfree tier on/offon
Base $430/moPrice → $49 $612/mo ×No freemium, $39 $705/mo ×Cut COGS, hold $19 $286/mo ×
PROJECTED MRR · +90D $430 vs base — Save as scenario Reset to base
The agent link

Pair once. It takes
tasking from the plan.

A one-time device-code pairing over MCP. After that, your agent pulls approved briefs straight from the launch plan and pushes status back. Scoped to one product + repo, revocable anytime — your keys never leave your box.

capture_product_contextv3.1
apply_fixv2.9
sync_planv3.0
pair your agent
G7Q4··

Enter this code in your agent, or run
npx @gtm/mcp pair

waiting for claude-code · scoped to 1 repo

Ship the thing. We'll show you exactly what's left.

Founding users get their first full review free on their live product — leave your email to claim a slot.

Sign up for launch notification