mcp-server-trello-growth

The growth ledger for @delorenj/mcp-server-trello. The maintainer made one strategic choice explicitly: distribution first — grow reach before revenue. This skill records that near-term reach work (Phase 1, do now) and parks the monetization thesis as a deferred, written-down bet (Phase 2, recorded, not built). It exists so the money idea is captured without being acted on, and so the reach work is gated and verifiable rather than vibes.

This skill runs on the same epistemic contract as the rest of the house. Every factual claim carries exactly one marker: VERIFIED (a command was run this session and the output cited), BELIEVED (plausible, unchecked — may never justify a spend on its own), or FALSIFIED. The "VERIFIED" facts below were true on 2026-07-21; the maintainer pivots fast — re-run the commands before you lean on any of them. llr is the recency compass.

---

What changed on 2026-07-21 (why reach work is worth doing now)

VERIFIED this session. After a 7-month release gap (npm sat on 1.7.1 from 2025-12-18), 1.8.0 shipped to both registries:

• npm latest = 1.8.0, time.modified = 2026-07-21T03:08Z, published via Trusted Publishing / OIDC (tokenless) — npm view @delorenj/mcp-server-trello version dist-tags.latest time.modified
• MCP registry io.github.delorenj/mcp-server-trello @ 1.8.0, isLatest: true — curl "https://registry.modelcontextprotocol.io/v0/servers?search=mcp-server-trello"

This is the unlock for growth, for three reasons:

1. New installs finally run current code. For 7 months every new install ran December's binary; 30+ merged PRs were invisible. Distribution work was pointless while the package was stale — you'd be funnelling people to a frozen build. Now it isn't.
2. The tracker's contamination should start to clear. Issues that requested features which already existed (artifacts of the release gap) can now be closed as shipped, not chased as demand.
3. The funnel is worth widening only once the thing at the end works. Fixing a broken discovery channel that leads to a 7-month-old package is negative ROI. Post-1.8.0 it's positive.

Say this first in any growth conversation, until it's false: the release gap is closed. Before 2026-07-21, "grow reach" was premature. After it, it's the live priority.

---

Segment the tracker BEFORE inferring demand (non-negotiable)

Per TheGardner: the issue tracker is a contaminated instrument. A 7-month publish failure means an unknown fraction of open issues measure "the package was broken/stale," not "users want this." Do not read a raw issue count as demand.

• Segment pre/post-1.7.1 (the release-gap cohort), and now pre/post-1.8.0 (the just-shipped cohort). Complaints filed against a stale package are not evidence about the current one.
• Open ≠ unresolved. Live example (VERIFIED this session): #2 "NPM package doesn't exist" is still OPEN, yet npm view returns 1.8.0. The underlying claim is FALSIFIED — the package plainly exists. The issue is just unclosed, and counting it as demand would be double-counting a fixed bug.
• Never fabricate demand, download counts, or user quotes. Cite behavior (a third party hand-assembling the loop, repeated PRs, a daily user's issue) or cite Atlassian's documented gap — and say which. "Every dev wants this" is not a data point.

---

PHASE 1 — DISTRIBUTION (do now)

Base to grow from, VERIFIED this session:

Metric	Value	Command
npm downloads, last 30d	8,882 (2026-06-21→07-20)	curl "https://api.npmjs.org/downloads/point/last-month/@delorenj/mcp-server-trello"
GitHub stars / forks	423 / 139	gh repo view delorenj/mcp-server-trello --json stargazerCount,forkCount
npm latest	1.8.0	npm view @delorenj/mcp-server-trello version

The often-quoted ~70k is cumulative downloads — a vanity number. The rate that actually reflects reach is ~8.9k/month (VERIFIED above; close to the ~9.3k BELIEVED in TheGardner 2026-07-16 — now re-confirmed). Track the rate, the stars trend, and — new post-1.8.0 — the share of installs on the current major. Do not celebrate the cumulative counter.

The reach moves — a gated ledger

Issue numbers come from TheGardner (dated 2026-07-16) and are BELIEVED there; their OPEN state was re-VERIFIED this session (gh issue view), but an open state is not proof the underlying problem persists (see #2 above). Re-read each issue body before acting.

#	Move	Why it grows reach	First action	Gate / risk	How to VERIFY it's actually fixed
1	Smithery listing (#47)	A broken discovery channel is negative reach — it advertises a dead install. Highest-priority funnel fix.	Reproduce the documented Smithery install end-to-end; find where it breaks (build config, smithery.yaml/metadata, or a stale version pin).	Don't "fix" blind — the break mechanism is BELIEVED, unconfirmed. Confirm the failure first.	Run the Smithery install flow to a booted server that lists the 1.8.0 tool count over stdio. Not "the page loads" — the install must succeed.
2	Claude marketplace / directory (#70)	A large and growing share of MCP demand now originates in-client. Absent = invisible to that whole channel. BELIEVED (share unquantified) — but low-regret to list.	Submit/enable the directory listing per Anthropic's current process; point it at 1.8.0.	Listing metadata drifts from the package — a listing pinned to an old version re-creates the stale-install problem.	Confirm the entry resolves and its advertised version == npm view … version (1.8.0). Re-check on each release.
3	Trello Power-Up marketplace (#39)	Trello-native discovery — a channel npm-only competitors structurally don't have. Meets users where the board already is.	Scope a Power-Up entry via the Trello developer portal; decide what it links to (install/docs), since the server is stdio, not a hosted Power-Up.	Highest effort of the five; needs a Trello dev-portal submission + review. Don't start until #47/#70 (cheaper) are done.	Find the Power-Up in Trello's in-product directory search and complete an install→docs handoff.
4	Onboarding funnel (#21)	key+token is a strictly worse funnel than OAuth consent. TheGardner: "the moat's benefit is claimed, its cost never booked." Every extra manual step is top-of-funnel leakage. BELIEVED the friction costs conversions — measure it.	Time a cold-start: fresh user → first successful tool call. Count steps and dead-ends (where does #21's "stuck at auth" happen?).	The locality moat depends on local key+token (no cloud OAuth round-trip). Reducing friction ≠ abandoning it — improve the docs/first-run path, don't propose cloud OAuth here.	Re-time the cold-start after the change; step count and time-to-first-call must drop. Watch for new #21-class "stuck at auth" reports.
5	npm discoverability (#2, historical)	Keywords + an above-the-fold README pitch decide whether npm search surfaces you and whether a visitor installs. Cheapest lever of all.	Add the "why this over the official Atlassian one" pitch above the fold (judgment/workflow encapsulation, not endpoint count). Tighten keywords.	#2's literal complaint is FALSIFIED (package exists). Don't reopen a fixed bug; treat #2 as a discoverability prompt, not a live defect.	npm view … keywords description reflects the new pitch; the README's first screen answers "why not the official one?"; search "trello mcp" and note the rank moved.

Where each move stands, grounded

• npm keywords today (VERIFIED, npm view … keywords description): mcp, trello, bun, automation, productivity, model-context-protocol, api, integration, typescript, board-management, task-management, kanban. The description is "An MCP server for Trello boards, powered by Bun for maximum performance." — it sells speed, not judgment. The moat is opinionated workflow encapsulation; the above-the-fold pitch should say so. (Move 5.)
• Competitive framing for Move 5 (BELIEVED, from TheGardner): we're the incumbent — 422★ vs 35/34/3 for the nearest 1:1-mapping servers; Composio ships ~329 tools (16× past the selection cliff). The official Trello MCP (mcp.trello.com/v1) is a cloud OAuth endpoint with no filesystem, cwd, or git — the locality gap is the pitch. Re-verify star/tool counts before quoting them in public copy.

---

PHASE 2 — MONETIZATION THESIS (recorded, NOT built)

This is a parked hypothesis with an explicit build-trigger. Under the distribution-first choice, nothing in this phase gets built or even instrumented yet. It is written down so it isn't lost, and gated so it isn't acted on prematurely.

The open-core seam

Read the locality moat from the money side and it splits cleanly:

• OSS (MIT) — the local writer + local interpreter. git, cwd, filesystem, private, no OAuth round-trip. This is the moat nobody can take — a cloud OAuth server (mcp.trello.com/v1) structurally cannot reach a filesystem or a repo. Keep this MIT. Forever.
• PAID — the layer stdio structurally cannot be. State that outlives a session; a readable back-channel that finally closes the inert watch_* loop (today watch_card/watch_list set subscribed:true into a Trello inbox this server cannot read — the write half of a loop with no read half — and "stdio has nowhere to put the cursor"); a team-shared corpus; history beyond the stdio window; hosted flow dashboards. Call it "the server that remembers."

Why it's not a rug-pull

You are not fencing off code that was free. The server stays MIT. The paid thing is a separate product selling a topology the OSS never had — persistence, a read-back channel, shared state, hosting. That is the same locality moat read from the money side: OSS owns "local, private, yours"; paid owns "durable, shared, hosted." Selling a capability the free version never possessed is not a bait-and-switch.

The build-trigger (the gate — do NOT cross it yet)

Do not build or instrument Phase 2 until both hold:

1. Reach is materially larger than the 2026-07-21 base (~8.9k downloads/mo, 423★). Distribution first means this comes before any monetization signal.
2. Demand is MEASURED, not assumed — via low-regret instruments, not a built product.

Future low-regret instruments (listed, explicitly NOT to be set up now):

• GitHub Sponsors — the lowest-regret willingness-to-pay signal. (VERIFIED this session: no .github/FUNDING.yml exists — Sponsors is not configured, and under distribution-first it stays that way for now.)
• A Team/Pro waitlist as a real instrument — a landing page that measures intent, not a product. A signup is a behavioral signal; a survey answer is not.

Rule: no Sponsors, no waitlist, no FUNDING.yml, no pricing page until Phase 1 has moved the reach numbers AND one of these instruments shows real signal. Recording the thesis is this skill's job; building it is a later decision.

The precedent that de-risks it (BELIEVED — from TheGardner, recheck)

Corrello and Nave already CHARGE for Trello board analytics as Power-Ups — cycle time, WIP age, dwell distributions. That is TheGardner's #1 vein ("interpretation of the transition log") with proven willingness-to-pay in the exact market. It is BELIEVED-from-TheGardner (dated 2026-07-16) — confirm both products still exist and still charge before citing this to justify a build. Note the shape: they sell interpretation of ambient data, which is precisely what stdio-plus-persistence ("the server that remembers") would own and what Trello ships natively nowhere.

---

Anti-patterns (each would burn reach or violate the choice)

• Funnelling installs to a stale package. The pre-2026-07-21 mistake. Never promote a discovery channel whose endpoint is an old build — verify npm view … version == the intended major first.
• Inferring demand from a raw issue count. Segment pre/post-1.7.1 and pre/post-1.8.0 first. #2 is the standing proof an open issue can be a fixed bug.
• Quoting the ~70k cumulative number as growth. It's a vanity total; track the ~8.9k/mo rate and the star trend.
• "Fixing" #47 blind. The Smithery break mechanism is BELIEVED — reproduce it before changing config, or you fix the wrong thing and still ship a dead install.
• Building Phase 2 (or setting up Sponsors / a waitlist / FUNDING.yml) now. The maintainer chose distribution first. Phase 2 is recorded, gated, and untouched until the build-trigger fires.
• Framing the paid product as gating the OSS. It sells a topology the OSS never had; keep the server MIT and say so, or the "not a rug-pull" claim collapses.