The site maps partners.json category directly to the sponsorship-page
category grid (CATEGORIES[].name in partner-program.ts), where the
canonical slot is "Website & landing page builders" — "Website" would
never match. Display surfaces are unaffected (they use categoryShort;
sync confirms README/REGISTRY unchanged).
Claude-Session: https://claude.ai/code/session_01VTdwoFxvAL9yNhkuH1EdpS
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* feat: publish Ploy as a live Skill Partner (Website)
Ploy's payment cleared (agreement 2026-08, accepted 2026-09-03), so the
staged entry goes live. The integration guide was finalized against
docs.ploy.ai (llms.txt + CLI/auth/webhooks docs) rather than waiting on
the onboarding packet — the maintainer holds the pen either way.
- partners.json: ploy active:true
- README + REGISTRY Verified Partners table: regenerated via sync-partners.mjs
- REGISTRY Tool Index: add the ◆ ploy capability-matrix row (Website)
- tools/integrations/ploy.md: disclosed ◆ guide — CLI, WebMCP, API tokens,
webhooks, PloyDB/Ploybooks, and how it fits the website-build/seo skills
- .partners-pending/ploy.md draft removed (superseded, was gitignored)
Logo ships separately in the marketingskills-website repo (/logos/ploy.png).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VTdwoFxvAL9yNhkuH1EdpS
* review: fix dangling skill references in ploy guide
website-build-* and ai-integration live in other skill libraries, not this
repo — the same defect class 5339aad fixed. Reference only skills that
exist here; add schema to the SEO pairing.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VTdwoFxvAL9yNhkuH1EdpS
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Ploy (ploy.ai) completed Skill Partner checkout (agreement 2026-08 accepted
2026-09-03; $500/mo). Payment has cleared but the onboarding packet (logo,
blurb, docs) hasn't arrived yet, so the entry is staged inactive — the
Converly #571 pattern in reverse order.
- partners.json: ploy entry, category Website (their slot per Corey),
active:false so no surface publishes yet
- draft integration guide staged locally in .partners-pending/ploy.md
(gitignored) with TODOs to fill from the onboarding packet
To go live once the packet arrives: finalize the guide into
tools/integrations/ploy.md, flip active:true, run
node scripts/sync-partners.mjs, add the ◆ rows + Website category to
tools/REGISTRY.md.
Claude-Session: https://claude.ai/code/session_01VTdwoFxvAL9yNhkuH1EdpS
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* feat: onboard Converly as a paying Skill Partner
Aaron Beashel completed payment and the onboarding form, so Converly goes
live — the inverse of the #571 unpublish.
- partners.json: active:true
- README + REGISTRY Verified Partners table: regenerated via sync-partners.mjs
- REGISTRY Tool Index: restore the ◆ converly capability-matrix row
- tools/integrations/converly.md: restored from .partners-pending/
The disclosure header ("Converly sponsors Marketing Skills") is now accurate.
Site picks this up via the live partners.json fetch once merged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* feat: set Converly logoUrl to self-hosted placement asset
Points the live-fetch path at the logo grabbed from converly.io and
self-hosted on the site (/logos/converly.png), so the Skill Partner card
shows the C mark instead of the fallback monogram.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Converly was used as the example partner to build out the partner pipeline
and went live across the README, registry, and site before any invoice was
paid. No company should appear as a paying partner before payment clears.
- partners.json: active:false (record kept for history)
- README + tools/REGISTRY.md: regenerated via sync — now "no active partners"
- tools/integrations/converly.md: moved to gitignored .partners-pending/
Restore on payment: flip active:true, restore the guide from .partners-pending/,
run node scripts/sync-partners.mjs.
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Consolidates the previously-scattered rules into one governance doc: the
one principle (sponsorship funds work, never recommendations), the three
tool-entry kinds (neutral integration / ◆ Verified Partner / house tool),
what sponsorship buys and never buys, tiers, the file-level disclosure
format, editorial control + the integrity rubric, the precise category-
exclusivity definition (slot not recommendations), the full partner
lifecycle, and the tools/integrations spec. Nothing new invented —
consolidated from the site FAQs, REGISTRY, CONTRIBUTING, and the
agreement. Cross-linked from REGISTRY, CONTRIBUTING, and the README
Partners section.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Codifies the editorial policy: anyone (tool makers, partners included) may
contribute to any skill; the bar is whether the content makes agents *aware*
of a tool without manufacturing a recommendation. Adds the rubric (options
not one answer, at the point of relevance, no forced endorsement, facts over
framing, disclose self-interest, the swap test, maintainer final review) to
CONTRIBUTING.md, a checklist item, and a pointer from the tools registry.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
partners.json is now the canonical partner registry. scripts/sync-partners.mjs
regenerates the README Partners section and the tools/REGISTRY.md Verified
Partners table (between PARTNERS:START/END markers) from it — run after editing
partners.json, or 'node scripts/sync-partners.mjs --check' in CI. The
marketing-skills.com site reads the same partners.json directly (raw GitHub), so
one edit updates every surface. Deactivate a lapsed partner with active:false.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Architecture change per the partner UX: a partner is a vetted, disclosed
tool integration in the existing registry — surfaced through the tool
routing core skills already use — not a nested skills/partners/ skill
(which wouldn't install and duplicated the tools layer).
- tools/integrations/converly.md: partner-grade integration doc with a
disclosure header (◆ Verified Partner; additive, never biasing)
- tools/REGISTRY.md: new Verified Partners section (what the ◆ marker
means + doesn't) + Converly row marked ◆ in the index
- README: Partners section now points at the registry/integration
- removed skills/partners/ draft (skill + convention doc) — the
convention now lives in the registry's Verified Partners section
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds .github/workflows/release.yml — on push to main, if the plugin.json
version changed and no v<version> tag exists, it creates the tag + a
GitHub Release with notes pulled from the matching VERSIONS.md block.
Idempotent (skips if already tagged). Closes the manual-release gap that
left v2.10.1–v2.11.0 untagged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- repo 2.10.4 -> 2.10.5 (plugin.json + marketplace.json)
- normalize content-strategy (2.1.1) and referrals (2.0.1) to patch bumps per AGENTS.md
- VERSIONS.md changelog for the 15-skill batch
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Enrich marketing-plan from Founding Marketing chs. 1 and 17.
Ch. 1 — new "Marketing as investing" north-star section in SKILL.md:
compounding portfolio ("no silver bullets, a hundred golden pellets"),
one-asset-many-returns, audition-not-an-auction, "hope is not a
strategy," plus the problem size x frequency market-quality gate as a
strategic gate in Section 2.
Ch. 17 — the two gaps not already excerpted in growth-patterns.md: the
70/20/10 resource-allocation rule (current / next / experimental) and a
weekly tracking cadence with plateau-indicator alerts that trigger the
next S-curve. Wired into the "how this informs the plan" table and
routed from SKILL.md.
Adds eval id 7. Bumps marketing-plan 1.1.0 -> 1.1.1 (metadata + VERSIONS).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Enrich content-strategy from Founding Marketing ch6 + ch14:
- ch6: "treat content like a product / brand surface area" thesis,
60/30/10 calendar split, and per-format execution discipline
(blog 10 titles/5 passes, guide, video, podcast, email). Clarify
the existing 40/30/20/10 scoring template.
- ch14: "Create Once, Distribute Twice" section — ORB as a
discovery->conversion funnel, creation-time distribution hooks,
the Content Distribution Flywheel, and failure modes
(spray-and-pray, platform dependency, ownership paradox).
- New references/content-distribution.md for the full distribution
spine (flywheel, platform half-lives, atomization checklist).
- Cross-links launch (ORB playbook) and social (atomization) rather
than re-deriving; wires SKILL.md routing.
- New eval (id 8); version 2.1.0 -> 2.2.0; VERSIONS row updated.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add a pre-launch "are you ready to launch?" readiness gate to the launch
skill, drawn from Founding Marketing ch.1. Runs before the five-stage
mechanics: uses Jason Cohen's SLC (Simple, Lovable, Complete) as the
alternative to a bare MVP, tied to the two chapter failure modes -
Stealth Mode (launching too late) and "Just One More Feature" (never
launching). SLC is the middle path.
- New SKILL.md "Readiness Gate" section before Five-Phase Launch Approach
- New eval (id 7) covering the gate + SLC + both failure modes
- Bump launch metadata.version 2.0.1 -> 2.0.2; update VERSIONS.md row
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add references/community-models.md adapting Founding Marketing Ch. 12:
the 5 named community models (Support-Driven/GreenPal,
Product-Development/Ydata, Education/LiveAgent, Founder-Led/Bento,
Postaga) with "best when" fit tests, the Notion ambassador benchmark
(300+ ambassadors, 1M+ template downloads, 25% of new users from
referrals), and the scaling-phase role shift (architect 0-100 ->
manager 100-1k -> enabler 1k+). Wire into SKILL.md routing, add eval
#7, bump to 2.0.1.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The skill covered the PR mix, pitch bar, and press page but had no
guidance on *what makes a story pitchable* when a founder has no news.
Adds the three earned-media angles from Founding Marketing ch.5 —
Founding Story, David vs Goliath, and Have an Enemy (a broken system,
never a competitor) — each with a when-to-use and exemplar, so the
skill can answer "how do we get press with no milestone?"
Also captures the compound-effect framing (press is for authority /
relationships / AI-citation surface, not the traffic bump), data
stories as a newsworthy-moment type, the "build media relationships
before you need them" ladder, and the one-asset PR repurposing
flywheel. New references/story-angles.md, wired into SKILL.md routing
and Core Philosophy, plus a first eval (id 1) covering angle selection.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds a named case library (Unsplash, HubSpot Website Grader, Baremetrics,
CoSchedule, Moz, Buffer, Shopify) with real conversion numbers, the
"your product is my marketing opportunity" framing, and a Common Pitfalls
list. Sourced from Founding Marketing ch. 8. Bumps free-tools to 2.0.1.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adapts chapter 3 of Corey Haines's Founding Marketing into the
copywriting skill (headline formulas already existed, so those are
skipped). Adds to references/copy-frameworks.md a new "Clarity &
Message-Market Fit" section covering the "Now you can" test, the Human
Action Model landing-page spine (discomfort -> vision -> path), the
Perception Gap with startup/enterprise value-prop swaps, the SavvyCal
awkward-link case (3x conversions), and the clarity metrics
(+81% conv, -38% cycle, -28% CAC, +175% referrals). SKILL.md routes to
it from Clarity Over Cleverness and the Above the Fold section. New
eval id 8. Bumps copywriting 2.0.1 -> 2.0.2 and its VERSIONS row.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The skill was mining/online-sleuthing heavy and thin on primary
research. Add an interviews + surveys layer so agents can help users
go ask customers directly, not just observe them in public.
New references/interviews-and-surveys.md covers the Sean Ellis /
Superhuman PMF survey (the "no longer use" question plus the 40% "very
disappointed" benchmark, Superhuman's 58%), Keep-Asking-Why 5-why
laddering with a churn-to-NRR worked example, a best-customers CRM
recruiting method, incentive guidance ($50/call, $5/survey; aim 10,
happy with 5), an outreach email template, Amy Hoy's Sales Safari,
and the confirmation-bias guardrail (prove yourself wrong; Dropbox).
Adapted from Founding Marketing ch.2.
SKILL.md gains a Mode 3 (primary research) that routes to the new
reference; description adds primary-research triggers; new eval id 13.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adapts Founding Marketing chs. 15-16 (pricing-rollout half) into the pricing skill:
- Initial "learning price" philosophy: pick a price you can learn from, the
$10/$100/$1,000 rule of thumb, the anti-$9 warning, and the Intercom/Jason
Fried "just charge $50 and see" story; steer off competitor/cost-based pricing
- references/pricing-models.md: the 8 pricing models (flat, usage, tier, user,
feature, credit, outcome, hybrid) with when-to-use + reference company, model
combining, the ~10:1 value/price ratio (Kulp), and the low-price-hurts-
retention counterpoint (Tringas/SpyFu)
- Pricing-change rollout methodology: test on new customers first, the anti-
grandfathering $50->$250 math, small-scale (5-10%) then staggered rollout,
communicate why months ahead with a generous offer
- New eval (id 8) for initial pricing + model selection
- Bump pricing 2.1.0 -> 2.1.1; update VERSIONS.md row (2026-08-23)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Give the ads skill a first-principles affordability gate for channel
decisions. LTV:CAC blesses channels that quietly run you out of cash: it
assumes universal churn, smooths over the front-loaded first-3-months
churn spike (Baremetrics), averages a $9 and a $999 plan into one
meaningless ratio, and ignores trial/sales-cycle revenue delay. Payback
Period (CAC/ARPU, target 3-12mo) and Discounted Payback
(CAC / (ARPU x retention)) are per-plan and time-aware, so the "can I
afford this channel?" question gets a real answer. Also captures two
adjacent stances from the source: OOH without social amplification is
wasted, and narrative-momentum ad copy tested one variable at a time.
Depth lives in references/payback-period.md (SKILL.md stays <500 lines);
wired into Reference Routing and cross-linked to attribution (CAC inputs)
and pricing (ARPU/plan structure). New eval (id 11) covers the
blended-LTV:CAC pushback + per-plan payback math.
Source: Corey Haines, Founding Marketing, ch. 7.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds references/partnership-types.md covering the five partnership types
from Founding Marketing Ch. 11 — integrations, reseller, affiliate,
co-marketing, and app store/marketplace — with flagship stats (HubSpot
$100M ARR / ~40% of revenue / 3,400+ partners) and standout moves
(Calendly's staged ladder, 20/80 affiliate power law, buyout clauses,
Notion's permissionless co-marketing, Grammarly 0→10M). Wires the
reference into SKILL.md routing with a summary table, adds eval id 7,
bumps co-marketing to 2.0.1 and updates the VERSIONS row.
Affiliate coverage stays at the partnership-strategy level; mechanics
remain in the referrals skill.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
