ProgrammableMarketing/mcpfy.ai (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 6
  • This quarter · 8
  • Big bets · 3
  • Wildcards · 3
  • How we can help
Generated Nov 4, 2026
  1. Programmable
  2. Playbooks
  3. Mcpfy
Mcpfy logo

growth playbook · Nov 4, 2026

mcpfy.ai (opens in a new tab)

Mcpfy growth playbook

Diagnosis

Mcpfy has a clear, timely product (one SDK and one cloud to get MCP servers and apps into ChatGPT, Claude and agents) and a fast, well-built marketing site. Early traction is small: 23 GitHub stars and about 45 npm installs a month for each package.

The growth problem is activation and trust. The docs start with the previous cURL-based product rather than the "npx create-mcpfy-app" flow the homepage sells, there are no guides for publishing to the stores, the Trust Center is still a placeholder, and homepage proof is hard to verify.

Alpic and Manufact sell the same promise, so a faster path to a live server and visible proof decide who wins.

Growth Score

6/10

Biggest Opportunity

Turn the homepage promise into a five-minute first deploy and own the "how to publish to ChatGPT and Claude" questions developers are asking now.

Biggest Risk

Better-funded rivals (Alpic, Manufact, Smithery under Arcade) target the same developers, and a weak first-run experience loses them to whoever gets them live first.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Make the first docs page match the homepage: "npx create-mcpfy-app" to a live URL in five minutes

    Make the SDK quick start the docs home page and move the cURL guide under "From an API".

    High impact First result in 2 weeks
  2. 2

    Replace the placeholder Trust Center before a Scale or Enterprise buyer reads it

    Write the hosting, secrets and incident-contact sections.

    High impact First result in 1 week
  3. 3

    Swap the unlinked social quotes for linked reviews and one named customer story

    Find the source of each quote.

    Medium impact First result in 2 weeks

02

Your idea portfolio

20 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 6 ideas are marked Unexpected: plays you are unlikely to have tried.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 3Big bets · 8Fill-ins · 7Reconsider · 2
Effort →

Showing 20 of 20 ideas

This week 6

Quick wins with a clear first action.

  • 01

    The homepage sells one SDK and one cloud, with "npx create-mcpfy-app" and "mcpfy dev" as the quick start. A developer who clicks "Docs" lands on "Create a Server", which describes the earlier product: find "Generate MCP" in the sidebar and paste cURL commands. The SDK getting-started page installs with npm install and never mentions create-mcpfy-app, and deploying from GitHub is a separate guide. The path from "this looks good" to a first deployed server breaks on the first page, and npm shows it: create-mcpfy-app had 42 downloads in the last 30 days.

    High impactLow effort First result in 2 weeks Growth Engineer

    First action

    Make the SDK quick start the docs home page and move the cURL guide under "From an API".

    How to measure it

    Sign-up to first deploy

    Now
    –
    Target
    40%
    By
    60 days

    Steps

    1. 1

      Write the quick start

      Scaffold, run, push, deploy, inspect, connect, with expected output for each step.

      Developer Advocate 2 days

    2. 2

      Reorder the docs

      Quick start first; cURL and OpenAPI under "From an API".

      Developer Advocate 1 day

    3. 3

      Match the homepage

      Same commands and wording on the homepage, SDK page and README.

      Growth Engineer 1 day

    4. 4

      Instrument it

      Events for scaffold, first deploy and first tool call.

      Founding Engineer 2 days

    5. 5

      Test it cold

      Five developers new to Mcpfy run it while you watch.

      Growth Engineer 1 week

    6. 6

      Fix the drop-offs

      Rewrite or automate the step where most people stop.

      Founding Engineer 1 week

    7. 7

      Report weekly

      Funnel from sign-up to first deploy.

      Growth Engineer weekly

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Growth Engineer

    Smallest test

    Swap the docs home page today and compare first deploys for two weeks.

    Why it works

    Replace the docs landing page with one quick start that matches the homepage: npx create-mcpfy-app, mcpfy dev, push to GitHub, get a live MCP URL, open it in the Cloud Inspector, then connect it to Claude or ChatGPT. Keep the OpenAPI and cURL route as a second path ("Already have an API?"). Put a time on each step, show the expected terminal output, and end with "Your server is live" plus a link to publishing. Track each step (scaffold, first deploy, first tool call in the Inspector) so you can see where people stop.

    For a developer tool, the first docs page is the onboarding. When it matches what the homepage promised, more visitors reach a live server, and a live server is the point where free users start to upgrade.

    Ready to use

    ChecklistFive-minute quick start

    [ ] npx create-mcpfy-app my-server
    [ ] cd my-server && mcpfy dev  (expect: ready on :4141, N tools registered)
    [ ] git push to a connected GitHub repo
    [ ] Copy the live MCP URL from the dashboard
    [ ] Call one tool in the Cloud Inspector
    [ ] Add the URL as a connector in Claude or ChatGPT
    [ ] Next: publishing checks and submission pack

    Risks and how to handle them

    • Breaking links to the old guide

      Keep the cURL page at its URL, linked from the new quick start.

    • Quick start drifts from the CLI

      Run it in CI on each SDK release.

    Evidence

    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
    • Your docsThe "Docs" link opens "Create a Server - MCPfy AI", the getting-started page, which describes the earlier flow: log in, "look for the 'Generate MCP' option in the sidebar", "Paste the cURL commands of the APIs that you need to add as tools", test and deploy. The SDK getting-started page installs with "npm install mcpfy-sdk" and never mentions create-mcpfy-app or mcpfy dev; deploy-from-GitHub is a separate guide.Source
    • Your docsThe docs index (llms.txt) lists 22 pages: Create a Server, Support, one GitHub deployment guide, 12 SDK pages (server, client, tools, prompts, resources, widgets, React widgets, authentication, Claude skill), 4 Pulse SDK telemetry pages for TypeScript and Python servers ("built on mcpfy-sdk or the official @modelcontextprotocol/sdk", "FastMCP or the official mcp SDK") and 3 API reference pages. There are no guides for publishing to the ChatGPT Apps store or Claude connectors, cross-client tests or the Inspector.Source
    • Open source and npmGitHub mcpfyy/mcpfy ("A SDK for building MCP tools, prompts, resources, and widgets", MIT, created July 2026): 23 stars, 8 forks, 7 open issues, last push September 15, 2026. Mcpfy/skills: 6 stars. npm last 30 days (to November 3, 2026): mcpfy-sdk 48 downloads, create-mcpfy-app 42; latest SDK 0.4.0 published September 15, 2026.Source
  • 02

    The Trust Center says what "a production trust center would document" and that "This demo makes no compliance claims". Scale ($299 a month) and Enterprise buyers run servers that touch their CRM and databases, and security review starts with that page.

    High impactLow effort First result in 1 week Founder and CEO

    First action

    Write the hosting, secrets and incident-contact sections.

    How to measure it

    Security questionnaires answered from the page

    Now
    –
    Target
    80%
    By
    90 days

    Steps

    1. 1

      List facts

      Hosting, encryption, secrets, access.

      Founding Engineer 1 day

    2. 2

      Subprocessors

      Cloud, email, analytics vendors.

      Founder and CEO 1 day

    3. 3

      Incident process

      Contact address and response time.

      Founder and CEO 1 day

    4. 4

      Write page

      Plain language, no placeholders.

      Founder and CEO 1 day

    5. 5

      Link it

      Pricing and Enterprise form.

      Growth Engineer 1 hour

    6. 6

      Roadmap

      Compliance plans with dates.

      Founder and CEO 1 week

    First result
    1 week
    Cost
    $0
    Owner
    Founder and CEO

    Smallest test

    Publish the hosting and secrets sections today.

    Why it works

    Write what is true today: where servers run, encryption in transit and at rest, how secrets are stored, access controls, logs, subprocessors, incident contact and response time, and any compliance work under way with dates. Link it from pricing and the Enterprise contact form.

    Plain, honest facts unblock security reviews; a placeholder ends them.

    Risks and how to handle them

    • Overclaiming

      Only state what is in place today.

    Evidence

    • Your websiteTrust Center page: "A production trust center would document encryption in transit and at rest, access controls, secret handling, and incident response" and "This demo makes no compliance claims. A real platform would list its certifications and reports here."Source
    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
  • 03

    The homepage quotes are credited to social handles but do not link to the original posts, and the screenshots use sample names. Developers check proof quickly, and quotes they cannot click weaken a strong product page.

    Medium impactLow effort First result in 2 weeks Growth Engineer

    First action

    Find the source of each quote.

    How to measure it

    Homepage sign-up rate

    Now
    –
    Target
    +10%
    By
    60 days

    Steps

    1. 1

      Source quotes

      Find original posts.

      Growth Engineer 1 day

    2. 2

      Replace

      Unlinked ones with linked reviews.

      Growth Engineer 1 day

    3. 3

      Ask a customer

      One short story.

      Founder and CEO 1 week

    4. 4

      Write it

      Problem, build, result.

      Developer Advocate 2 days

    5. 5

      Publish

      Homepage and blog.

      Growth Engineer 1 day

    6. 6

      Measure

      Sign-up rate.

      Growth Engineer monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Growth Engineer

    Smallest test

    Link the first three quotes today.

    Why it works

    Link each quote to its original post or replace it with a linked Product Hunt or GitHub comment, and add one short story from a logo-strip customer (what they built, time to deploy, clients used).

    Verifiable proof is what developer buyers trust.

    Risks and how to handle them

    • Customers slow to approve

      Start with a quote and logo.

    Evidence

    • Your websiteHomepage proof: a "Trusted by teams building on Mcpfy" logo strip (emergence, B Sabadell, Smile, ARGENISS, WebSeo, docfyn, creagia, DifferentByte, Zerone), a scrolling set of short quotes credited to social handles without links to the original posts, a Product Hunt badge, and product screenshots that use sample names ("Acme Labs", "Northwind CRM").Source
    • Open source and npmGitHub mcpfyy/mcpfy ("A SDK for building MCP tools, prompts, resources, and widgets", MIT, created July 2026): 23 stars, 8 forks, 7 open issues, last push September 15, 2026. Mcpfy/skills: 6 stars. npm last 30 days (to November 3, 2026): mcpfy-sdk 48 downloads, create-mcpfy-app 42; latest SDK 0.4.0 published September 15, 2026.Source
  • 04

    Mcpfy's big promise is distribution ("In the ChatGPT Apps Store. In Claude Connectors. From one build"), and Builder is pitched "For side projects heading to the marketplaces". The docs have no guide for either store, and no page on the submission pack or publishing checks.

    High impactMedium effort First result in 1 month Developer Advocate

    First action

    Outline the ChatGPT Apps store guide.

    How to measure it

    Guide-sourced sign-ups

    Now
    –
    Target
    100 a month
    By
    90 days

    Steps

    1. 1

      Research rules

      Current store requirements.

      Developer Advocate 3 days

    2. 2

      Write ChatGPT guide

      Full steps.

      Developer Advocate 1 week

    3. 3

      Write Claude guide

      Full steps.

      Developer Advocate 1 week

    4. 4

      Screenshots

      Real flows.

      Developer Advocate 2 days

    5. 5

      Publish

      Docs and blog.

      Growth Engineer 1 day

    6. 6

      Update

      When rules change.

      Developer Advocate monthly

    First result
    1 month
    Cost
    Writing time
    Owner
    Developer Advocate

    Smallest test

    Publish the ChatGPT guide first.

    Why it works

    Publish step-by-step guides for each store: requirements, review rules, icons and screenshots, common rejections, and how the publishing checks and submission pack handle them. Make them useful without Mcpfy, with Mcpfy as the shortcut.

    Developers search for these steps now; the guides bring them in at the moment they need the product.

    Risks and how to handle them

    • Rules change fast

      Date each guide and review monthly.

    Evidence

    • Your docsThe docs index (llms.txt) lists 22 pages: Create a Server, Support, one GitHub deployment guide, 12 SDK pages (server, client, tools, prompts, resources, widgets, React widgets, authentication, Claude skill), 4 Pulse SDK telemetry pages for TypeScript and Python servers ("built on mcpfy-sdk or the official @modelcontextprotocol/sdk", "FastMCP or the official mcp SDK") and 3 API reference pages. There are no guides for publishing to the ChatGPT Apps store or Claude connectors, cross-client tests or the Inspector.Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
    • Competitor researchRivals' own homepages: Alpic, "build, deploy, monitor and distribute your ChatGPT Apps and MCP servers"; Manufact (mcp-use.com), "The cloud for production MCP ... Get on ChatGPT and Claude plugin stores ... test every release across clients"; Smithery, "MCP servers, now part of Arcade"; Glama, "The MCP Server Registry, Inspector & Gateway".Source
  • 05

    The homepage offers templates (Chart Builder, Diagram Builder) and Git imports, but a new user still needs a repo, the CLI and a GitHub connection before seeing a live server.

    Medium impactLow effort First result in 1 month Founding Engineer

    First action

    Build the one-click flow for Chart Builder.

    How to measure it

    One-click deploys

    Now
    –
    Target
    50 a week
    By
    60 days

    Steps

    1. 1

      Pick template

      Chart Builder first.

      Founding Engineer 1 day

    2. 2

      Fork and deploy

      GitHub app flow.

      Founding Engineer 1 week

    3. 3

      Land in Inspector

      Tool pre-filled.

      Founding Engineer 2 days

    4. 4

      Add buttons

      Homepage and docs.

      Growth Engineer 1 day

    5. 5

      Track

      Deploy to upgrade.

      Growth Engineer weekly

    6. 6

      More templates

      Diagram and Postgres.

      Founding Engineer 1 month

    First result
    1 month
    Cost
    Engineering time
    Owner
    Founding Engineer

    Smallest test

    Ship one template button.

    Why it works

    Add "Deploy this template" buttons that fork the template into the user's GitHub and deploy it in one step, landing them in the Cloud Inspector with a tool ready to call.

    Seeing a live server in under a minute makes the product real before any code is written.

    Risks and how to handle them

    • Abandoned forks

      Clean up inactive projects on Free.

    Evidence

    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
  • 06

    PageSpeed flags one accessibility issue (low colour contrast) on an otherwise fast site, and the only blog post is titled "What is Mcpfy AI and Why Were Building IT".

    Low impactLow effort First result in 1 day Growth Engineer

    First action

    Fix both today.

    How to measure it

    Accessibility score

    Now
    –
    Target
    100
    By
    1 week

    Steps

    1. 1

      Find element

      From the PageSpeed report.

      Growth Engineer 15 minutes

    2. 2

      Fix colour

      Adjust the token.

      Growth Engineer 15 minutes

    3. 3

      Fix title

      And the URL redirect if changed.

      Growth Engineer 15 minutes

    4. 4

      Re-run

      PageSpeed.

      Growth Engineer 5 minutes

    5. 5

      Check others

      Platform pages.

      Growth Engineer 1 hour

    6. 6

      Add to checklist

      Before each launch.

      Growth Engineer 15 minutes

    First result
    1 day
    Cost
    $0
    Owner
    Growth Engineer

    Smallest test

    Fix the title now.

    Why it works

    Raise the contrast on the flagged text and correct the title to "What is Mcpfy and why we're building it".

    Small polish matters on a site that sells reliability to developers.

    Risks and how to handle them

    • Redirect breaks shares

      Keep the old URL redirecting.

    Evidence

    • Google PageSpeedwww.mcpfy.ai scores 96 on mobile and 100 on desktop, SEO 100 and accessibility 96, with one issue: "Background and foreground colors do not have a sufficient contrast ratio." Real visitors see the main content in 1.7 s at the 75th percentile.Source
    • Your websiteSitemap: 17 URLs, including 7 platform pages (hosting, cross-client testing, publishing checks, Cloud Inspector, public chat, analytics, submission pack), SDK, Inspector and Vibe pages, and 1 blog post ("What is Mcpfy AI and Why Were Building IT", September 12, 2026), which says "More than 24,700 MCP server repositories have already been created, and that number grew 55% last quarter".Source

This quarter 8

The core plays for the next three months.

  • 07

    Pulse SDK adds tool-level telemetry to servers built with the official MCP SDKs or FastMCP, in TypeScript and Python. That reaches far more developers than those who will move hosting, but the homepage barely mentions it.

    High impactMedium effort First result in 1 quarter Founder and CEO

    First action

    Write the Pulse landing page.

    How to measure it

    Pulse-connected servers

    Now
    –
    Target
    300
    By
    2 quarters

    Steps

    1. 1

      Landing page

      Two-line install.

      Growth Engineer 1 week

    2. 2

      Free tier

      Dashboard limits.

      Founder and CEO 1 week

    3. 3

      Python launch

      FastMCP community post.

      Developer Advocate 1 week

    4. 4

      TS launch

      Official SDK users.

      Developer Advocate 1 week

    5. 5

      Upsell path

      Hosting prompts in dashboard.

      Founding Engineer 2 weeks

    6. 6

      Measure

      Pulse to hosting.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Staff time
    Owner
    Founder and CEO

    Smallest test

    Post the Pulse install to one MCP community.

    Why it works

    Position Pulse as a free, two-line install for any MCP server, with its own landing page, a Python and TypeScript snippet, and a free dashboard. Offer hosting once users see their traffic.

    Observability is an easy first yes; hosting follows once the data lives in Mcpfy.

    Risks and how to handle them

    • Free users never upgrade

      Cap retention on free telemetry.

    Evidence

    • Your docsThe docs index (llms.txt) lists 22 pages: Create a Server, Support, one GitHub deployment guide, 12 SDK pages (server, client, tools, prompts, resources, widgets, React widgets, authentication, Claude skill), 4 Pulse SDK telemetry pages for TypeScript and Python servers ("built on mcpfy-sdk or the official @modelcontextprotocol/sdk", "FastMCP or the official mcp SDK") and 3 API reference pages. There are no guides for publishing to the ChatGPT Apps store or Claude connectors, cross-client tests or the Inspector.Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Your websiteSitemap: 17 URLs, including 7 platform pages (hosting, cross-client testing, publishing checks, Cloud Inspector, public chat, analytics, submission pack), SDK, Inspector and Vibe pages, and 1 blog post ("What is Mcpfy AI and Why Were Building IT", September 12, 2026), which says "More than 24,700 MCP server repositories have already been created, and that number grew 55% last quarter".Source
  • 08

    Mcpfy runs one suite "across clients and models, judged automatically" and claims servers are verified across ChatGPT, Claude, Gemini and Cursor. Developers struggle to know which features work in which client.

    High impactMedium effort First result in 1 quarter Developer Advocate

    First action

    Pick the first 20 servers.

    How to measure it

    Maintainer sign-ups

    Now
    –
    Target
    30
    By
    1 quarter

    Steps

    1. 1

      Pick servers

      Popular open-source ones.

      Developer Advocate 1 week

    2. 2

      Run suite

      Across clients.

      Founding Engineer 2 weeks

    3. 3

      Publish

      Table page.

      Growth Engineer 1 week

    4. 4

      Invite

      Maintainers.

      Developer Advocate 2 weeks

    5. 5

      Weekly refresh

      Automated.

      Founding Engineer weekly

    6. 6

      Measure

      Links and sign-ups.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Compute
    Owner
    Developer Advocate

    Smallest test

    Publish the table for 20 servers.

    Why it works

    Run the suite on 50 popular open-source MCP servers and publish a table: which tools and widgets work in each client, updated weekly. Invite maintainers to claim their entry.

    A useful public resource earns links and puts the testing product in front of maintainers.

    Risks and how to handle them

    • Maintainers dislike failing grades

      Show "works" and "known issues", not scores.

    Evidence

    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Your websiteSitemap: 17 URLs, including 7 platform pages (hosting, cross-client testing, publishing checks, Cloud Inspector, public chat, analytics, submission pack), SDK, Inspector and Vibe pages, and 1 blog post ("What is Mcpfy AI and Why Were Building IT", September 12, 2026), which says "More than 24,700 MCP server repositories have already been created, and that number grew 55% last quarter".Source
    • Competitor researchRivals' own homepages: Alpic, "build, deploy, monitor and distribute your ChatGPT Apps and MCP servers"; Manufact (mcp-use.com), "The cloud for production MCP ... Get on ChatGPT and Claude plugin stores ... test every release across clients"; Smithery, "MCP servers, now part of Arcade"; Glama, "The MCP Server Registry, Inspector & Gateway".Source
  • 09

    The mcpfy SDK is MIT licensed with 23 stars, 8 forks and 7 open issues, and the last push was in mid-September. Developer tools grow from repos that look alive.

    Medium impactMedium effort First result in 1 quarter Developer Advocate

    First action

    Close or answer the seven open issues.

    How to measure it

    SDK stars

    Now
    –
    Target
    500
    By
    2 quarters

    Steps

    1. 1

      Triage issues

      All open ones.

      Founding Engineer 2 days

    2. 2

      Release cadence

      Monthly changelog.

      Founding Engineer monthly

    3. 3

      Examples

      Five servers.

      Developer Advocate 1 month

    4. 4

      Build-alongs

      Posts and videos.

      Developer Advocate monthly

    5. 5

      Share

      Developer communities.

      Developer Advocate monthly

    6. 6

      Track

      Stars, forks, deploys.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Developer Advocate

    Smallest test

    Publish one example this week.

    Why it works

    Answer open issues, ship monthly releases, add five example servers people need (Postgres, Notion-style docs, charts, Stripe-style billing, internal API), and post each as a short build-along.

    Active repos with copyable examples get starred, forked and deployed.

    Risks and how to handle them

    • Spreading too thin

      Pick examples from real user requests.

    Evidence

    • Open source and npmGitHub mcpfyy/mcpfy ("A SDK for building MCP tools, prompts, resources, and widgets", MIT, created July 2026): 23 stars, 8 forks, 7 open issues, last push September 15, 2026. Mcpfy/skills: 6 stars. npm last 30 days (to November 3, 2026): mcpfy-sdk 48 downloads, create-mcpfy-app 42; latest SDK 0.4.0 published September 15, 2026.Source
    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
  • 10

    The blog has one post, from September. Mcpfy's own post says MCP server repositories grew 55% last quarter; developers are searching for how to build widgets, handle auth and debug tool calls.

    High impactLow effort First result in 1 quarter Developer Advocate

    First action

    List the ten questions users ask most.

    How to measure it

    Blog-sourced sign-ups

    Now
    –
    Target
    150 a month
    By
    2 quarters

    Steps

    1. 1

      Question list

      From support and Slack.

      Developer Advocate 1 day

    2. 2

      Calendar

      Ten posts.

      Developer Advocate 1 day

    3. 3

      Write

      Weekly.

      Developer Advocate weekly

    4. 4

      Deploy buttons

      In each post.

      Growth Engineer weekly

    5. 5

      Share

      Communities.

      Developer Advocate weekly

    6. 6

      Measure

      Sign-ups.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Writing time
    Owner
    Developer Advocate

    Smallest test

    Publish the first post this week.

    Why it works

    Publish one practical post a week (widget patterns, OAuth for MCP, debugging JSON-RPC, cold starts) with code and a deploy button.

    Steady technical content compounds in search and in AI assistants that answer developer questions.

    Risks and how to handle them

    • Posts too generic

      Each one solves one specific problem.

    Evidence

    • Your websiteSitemap: 17 URLs, including 7 platform pages (hosting, cross-client testing, publishing checks, Cloud Inspector, public chat, analytics, submission pack), SDK, Inspector and Vibe pages, and 1 blog post ("What is Mcpfy AI and Why Were Building IT", September 12, 2026), which says "More than 24,700 MCP server repositories have already been created, and that number grew 55% last quarter".Source
    • Your docsThe docs index (llms.txt) lists 22 pages: Create a Server, Support, one GitHub deployment guide, 12 SDK pages (server, client, tools, prompts, resources, widgets, React widgets, authentication, Claude skill), 4 Pulse SDK telemetry pages for TypeScript and Python servers ("built on mcpfy-sdk or the official @modelcontextprotocol/sdk", "FastMCP or the official mcp SDK") and 3 API reference pages. There are no guides for publishing to the ChatGPT Apps store or Claude connectors, cross-client tests or the Inspector.Source
  • 11

    Alpic and Manufact now pitch the same thing (build, deploy, test across clients and get into the ChatGPT and Claude stores), and Glama offers a registry, Inspector and gateway. Developers compare them side by side.

    Medium impactLow effort First result in 1 month Founder and CEO

    First action

    Write "Mcpfy vs Alpic" first.

    How to measure it

    Comparison sign-ups

    Now
    –
    Target
    30 a month
    By
    90 days

    Steps

    1. 1

      Research rivals

      Their own pages and pricing.

      Founder and CEO 3 days

    2. 2

      Write

      Three pages.

      Developer Advocate 1 week

    3. 3

      Review

      Accuracy.

      Founder and CEO 1 day

    4. 4

      Publish

      With dates.

      Growth Engineer 1 day

    5. 5

      Link

      From pricing.

      Growth Engineer 1 hour

    6. 6

      Refresh

      Quarterly.

      Developer Advocate quarterly

    First result
    1 month
    Cost
    $0
    Owner
    Founder and CEO

    Smallest test

    Publish one comparison.

    Why it works

    Write fair comparisons: SDK, hosting, preview URLs, testing, analytics, store publishing, pricing. Be clear where each fits.

    Comparison pages capture late-stage searches and sharpen the pitch.

    Risks and how to handle them

    • Fast-moving rivals

      Date every claim.

    Evidence

    • Competitor researchRivals' own homepages: Alpic, "build, deploy, monitor and distribute your ChatGPT Apps and MCP servers"; Manufact (mcp-use.com), "The cloud for production MCP ... Get on ChatGPT and Claude plugin stores ... test every release across clients"; Smithery, "MCP servers, now part of Arcade"; Glama, "The MCP Server Registry, Inspector & Gateway".Source
    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
  • 12

    Free has 2 projects and 7-day logs; Builder adds projects, 30-day logs, previews and cold-start prevention for $29. Users feel these limits at specific moments.

    Medium impactMedium effort First result in 1 quarter Growth Engineer

    First action

    Add the third-project prompt.

    How to measure it

    Free-to-paid rate

    Now
    –
    Target
    5%
    By
    1 quarter

    Steps

    1. 1

      Map moments

      Three limits.

      Growth Engineer 1 day

    2. 2

      Design prompts

      One line each.

      Growth Engineer 2 days

    3. 3

      Build

      In the dashboard.

      Founding Engineer 1 week

    4. 4

      Email follow-up

      If dismissed.

      Growth Engineer 2 days

    5. 5

      Test

      Against no prompt.

      Growth Engineer 1 month

    6. 6

      Tune

      Copy and timing.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Growth Engineer

    Smallest test

    Add the third-project prompt.

    Why it works

    Show an upgrade prompt exactly then: creating a third project, opening logs older than 7 days, or a cold-start delay on a public server, with the Builder benefit spelled out.

    Prompts at the moment of need convert far better than a pricing page.

    Risks and how to handle them

    • Feels pushy

      Show once per limit.

    Evidence

    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
  • 13

    Mcpfy started as "Turn APIs into Production-Ready MCP Servers" and still offers OpenAPI import. Many SaaS companies want an MCP server for their API without writing one.

    High impactMedium effort First result in 1 quarter Founding Engineer

    First action

    Build the paste-a-URL preview.

    How to measure it

    Preview-to-sign-up rate

    Now
    –
    Target
    25%
    By
    1 quarter

    Steps

    1. 1

      Preview flow

      Public spec only.

      Founding Engineer 2 weeks

    2. 2

      Tool list

      Shown in Inspector.

      Founding Engineer 1 week

    3. 3

      Landing page

      For SaaS teams.

      Growth Engineer 1 week

    4. 4

      Outreach

      APIs without MCP servers.

      Founder and CEO 1 month

    5. 5

      Follow-up

      Publishing help.

      Founder and CEO ongoing

    6. 6

      Measure

      Previews to deploys.

      Growth Engineer monthly

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Founding Engineer

    Smallest test

    Run the preview on ten public APIs.

    Why it works

    Build a landing page for SaaS teams: paste a public OpenAPI URL, get a preview server and a list of generated tools in the Inspector, then sign up to deploy and publish.

    A preview from their own API is a compelling first touch for teams that would never start from an SDK.

    Risks and how to handle them

    • Large specs produce too many tools

      Suggest a curated subset.

    Evidence

    • Change trackingprogrammable.info change tracking: a year ago the headline was "Turn APIs into Production-Ready MCP Servers"; now it is "Build and deploy MCP servers and apps". Google Analytics, Microsoft Clarity and Crisp chat were removed from the site in that time. The domain was registered in April 2025.Source
    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
  • 14

    Mcpfy can generate a server from an OpenAPI spec and publish it to ChatGPT and Claude. Many SaaS companies with public APIs have no MCP presence yet.

    High impactMedium effort First result in 2 quarters Founder and CEO

    First action

    Pick ten SaaS APIs with public specs.

    How to measure it

    Reply rate

    Now
    –
    Target
    20%
    By
    2 quarters

    Steps

    1. 1

      Find targets

      Public OpenAPI specs.

      Growth Engineer 1 week

    2. 2

      Generate

      Preview servers.

      Founding Engineer weekly

    3. 3

      Record

      Short clips.

      Developer Advocate weekly

    4. 4

      Send

      To DevRel or product.

      Founder and CEO weekly

    5. 5

      Follow up

      Offer to publish.

      Founder and CEO weekly

    6. 6

      Track

      Replies and deals.

      Growth Engineer monthly

    First result
    2 quarters
    Cost
    Staff time only
    Owner
    Founder and CEO

    Smallest test

    Send three previews this week.

    Why it works

    Each week, generate preview servers for ten SaaS APIs, record a short Inspector clip of their tools working in Claude, and send it to their developer relations or product lead.

    Showing a company its own product inside Claude is hard to ignore.

    Risks and how to handle them

    • Unwanted use of their API

      Keep previews private and share only with them.

    Evidence

    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Change trackingprogrammable.info change tracking: a year ago the headline was "Turn APIs into Production-Ready MCP Servers"; now it is "Build and deploy MCP servers and apps". Google Analytics, Microsoft Clarity and Crisp chat were removed from the site in that time. The domain was registered in April 2025.Source

Big bets 3

Larger plays that can change the trajectory.

  • 15

    Mcpfy already captures every tool call and helps publish to ChatGPT and Claude stores. Once apps are listed, owners will want installs, usage and ranking across stores in one view.

    High impactHigh effort First result in 3 quarters Founder and CEO

    First action

    Interview five teams with listed apps.

    How to measure it

    Scale retention at 6 months

    Now
    –
    Target
    90%
    By
    3 quarters

    Steps

    1. 1

      Interviews

      Five listed teams.

      Founder and CEO 2 weeks

    2. 2

      Design

      Dashboard.

      Founding Engineer 1 month

    3. 3

      Build

      First version.

      Founding Engineer 2 months

    4. 4

      Alerts

      Breakage detection.

      Founding Engineer 1 month

    5. 5

      Launch

      To Scale users.

      Growth Engineer 2 weeks

    6. 6

      Measure

      Retention.

      Growth Engineer quarterly

    First result
    3 quarters
    Cost
    Engineering time
    Owner
    Founder and CEO

    Smallest test

    Mock the dashboard and show five teams.

    Why it works

    Build a store dashboard: listing status, review feedback, installs and usage per client, with alerts when a client update breaks a tool.

    Owning the post-launch view keeps teams on Scale long after the first deploy.

    Risks and how to handle them

    • Store data access limited

      Start with Mcpfy-side usage.

    Evidence

    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
    • Competitor researchRivals' own homepages: Alpic, "build, deploy, monitor and distribute your ChatGPT Apps and MCP servers"; Manufact (mcp-use.com), "The cloud for production MCP ... Get on ChatGPT and Claude plugin stores ... test every release across clients"; Smithery, "MCP servers, now part of Arcade"; Glama, "The MCP Server Registry, Inspector & Gateway".Source
  • 16

    Several logo-strip names look like development agencies and consultancies. Agencies are being asked by clients to "get us into ChatGPT", and need a platform to build and hand over on.

    High impactHigh effort First result in 3 quarters Founder and CEO

    First action

    Ask two agency customers what they would need.

    How to measure it

    Paid projects from partners

    Now
    –
    Target
    50
    By
    3 quarters

    Steps

    1. 1

      Interview

      Two agencies.

      Founder and CEO 2 weeks

    2. 2

      Define plan

      Workspaces and share.

      Founder and CEO 1 month

    3. 3

      Build

      Multi-client support.

      Founding Engineer 2 months

    4. 4

      Directory

      Partner page.

      Growth Engineer 2 weeks

    5. 5

      Recruit

      Ten agencies.

      Founder and CEO 2 months

    6. 6

      Review

      Revenue per partner.

      Founder and CEO quarterly

    First result
    3 quarters
    Cost
    Engineering and partner time
    Owner
    Founder and CEO

    Smallest test

    Sign one agency to a pilot.

    Why it works

    Offer agencies a partner plan: multi-client workspaces, white-label handover, revenue share and a directory listing. Co-market the first case studies.

    Each agency brings many end clients on paid plans.

    Risks and how to handle them

    • Support burden

      Partner-first support tiers.

    Evidence

    • Your websiteHomepage proof: a "Trusted by teams building on Mcpfy" logo strip (emergence, B Sabadell, Smile, ARGENISS, WebSeo, docfyn, creagia, DifferentByte, Zerone), a scrolling set of short quotes credited to social handles without links to the original posts, a Product Hunt badge, and product screenshots that use sample names ("Acme Labs", "Northwind CRM").Source
    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
  • 17

    Mcpfy's first post already cites MCP repository growth. Its platform sees tool-call volume, latency, errors and client mix across hosted servers.

    High impactHigh effort First result in 2 quarters Developer Advocate

    First action

    Define five safe aggregate metrics.

    How to measure it

    Sign-ups per report

    Now
    –
    Target
    200
    By
    3 quarters

    Steps

    1. 1

      Metrics

      Five aggregates.

      Founding Engineer 1 week

    2. 2

      Privacy rules

      Minimum counts.

      Founder and CEO 1 week

    3. 3

      Pipeline

      Monthly export.

      Founding Engineer 2 weeks

    4. 4

      Write

      Report.

      Developer Advocate monthly

    5. 5

      Pitch

      Newsletters.

      Developer Advocate monthly

    6. 6

      Review

      After three.

      Founder and CEO quarterly

    First result
    2 quarters
    Cost
    Staff time
    Owner
    Developer Advocate

    Smallest test

    Publish one report on client share.

    Why it works

    Publish monthly anonymised numbers: client share (ChatGPT, Claude, Cursor), median latency, common errors and widget usage. Pitch it to developer newsletters.

    Original data makes Mcpfy the source people quote about MCP adoption.

    Risks and how to handle them

    • Small sample early

      Say how many servers it covers.

    Evidence

    • Your websiteSitemap: 17 URLs, including 7 platform pages (hosting, cross-client testing, publishing checks, Cloud Inspector, public chat, analytics, submission pack), SDK, Inspector and Vibe pages, and 1 blog post ("What is Mcpfy AI and Why Were Building IT", September 12, 2026), which says "More than 24,700 MCP server repositories have already been created, and that number grew 55% last quarter".Source
    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source

Wildcards 3

Unexpected experiments, each with a cheap way to test it.

  • 18

    Mcpfy has a "Public chat" feature. Visitors understand MCP apps fastest by using one.

    Medium impactLow effort First result in 1 month Developer Advocate

    First action

    Pick the demo app.

    How to measure it

    Demo-to-sign-up rate

    Now
    –
    Target
    10%
    By
    2 months

    Steps

    1. 1

      Pick app

      Chart demo.

      Founding Engineer 1 day

    2. 2

      Embed

      Homepage.

      Founding Engineer 1 week

    3. 3

      Limits

      Rate limits.

      Founding Engineer 2 days

    4. 4

      CTA

      After first answer.

      Growth Engineer 1 day

    5. 5

      Track

      Sign-ups.

      Growth Engineer weekly

    6. 6

      Rotate

      New demos monthly.

      Developer Advocate monthly

    First result
    1 month
    Cost
    Model costs
    Owner
    Developer Advocate

    Smallest test

    Share the public chat link in one community.

    Why it works

    Host a public chat on the homepage with a sample app (charts from a demo dataset), and show "Build this in five minutes" after the first answer.

    Trying it removes the abstraction around MCP for newcomers.

    Risks and how to handle them

    • Abuse

      Rate-limit and cap tokens.

    Evidence

    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source
    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
  • 19

    Builder is aimed at "side projects heading to the marketplaces". Hackathons draw exactly those builders.

    Medium impactLow effort First result in 1 month Developer Advocate

    First action

    Set the first theme and date.

    How to measure it

    Deployed projects

    Now
    –
    Target
    40 per event
    By
    1 quarter

    Steps

    1. 1

      Theme

      One per month.

      Developer Advocate 1 day

    2. 2

      Page

      Rules and sign-up.

      Growth Engineer 2 days

    3. 3

      Promote

      Communities.

      Developer Advocate 2 weeks

    4. 4

      Support

      Live help channel.

      Developer Advocate weekend

    5. 5

      Judge

      Working apps.

      Founder and CEO 1 day

    6. 6

      Feature

      Winners on blog.

      Developer Advocate 1 week

    First result
    1 month
    Cost
    Builder credits
    Owner
    Developer Advocate

    Smallest test

    Run one weekend.

    Why it works

    Run a monthly online build weekend with a theme, the quick start and templates, judged on a working app in Claude or ChatGPT; prizes are Builder months and a feature on the blog.

    It creates projects, content and word of mouth at low cost.

    Risks and how to handle them

    • Low turnout

      Partner with a developer community.

    Evidence

    • Your websitePricing: Free $0 (2 projects, 7-day logs), Builder $29 a month ("For side projects heading to the marketplaces", 5 projects, preview deployments, cold-start prevention), Scale $299 a month (unlimited projects, 1-year logs, all regions, dedicated Slack channel), Enterprise custom; yearly billing saves 17%.Source
    • Your websiteHomepage: "Build and deploy MCP servers and apps. One SDK, one cloud, every client. Ship to ChatGPT and Claude, run for your own agents, and see every tool call in production." Quick start shown as "npx create-mcpfy-app" then "mcpfy dev ready on :4141"; three ways in: from a Git repository, from an OpenAPI spec, or from a template; calls to action "Start deploying" and "Book a call".Source
  • 20

    Mcpfy runs tests across ChatGPT, Claude, Gemini and Cursor. Clients change often, and developers usually learn of breakage from their users.

    Medium impactLow effort First result in 1 month Founding Engineer

    First action

    Detect one client update and rerun suites.

    How to measure it

    Alert-to-login rate

    Now
    –
    Target
    40%
    By
    1 quarter

    Steps

    1. 1

      Detect updates

      Client release notes.

      Founding Engineer 1 week

    2. 2

      Rerun suites

      Automated.

      Founding Engineer 2 weeks

    3. 3

      Email summary

      Pass/fail.

      Growth Engineer 1 week

    4. 4

      Link

      To the failing call.

      Founding Engineer 1 week

    5. 5

      Upsell

      Deeper tests on Scale.

      Growth Engineer 1 week

    6. 6

      Measure

      Returns.

      Growth Engineer monthly

    First result
    1 month
    Cost
    Compute
    Owner
    Founding Engineer

    Smallest test

    Send one alert after the next client update.

    Why it works

    Rerun each user's cross-client suite when a major client update ships and email them a pass/fail summary, free on all plans.

    A useful alert brings users back to the dashboard and shows the value of paid testing.

    Risks and how to handle them

    • False alarms

      Retry before alerting.

    Evidence

    • Your websitePlatform claims: "Verified across ChatGPT Claude Gemini Cursor"; a live MCP URL for every branch; "Every JSON-RPC frame captured, with session replay"; "One suite, run across clients and models, judged automatically"; "Publish checks, then a generated submission pack" for connector directories and app stores; the Cloud Inspector runs "the same call across GPT, Claude, and Gemini".Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

ProgrammableInfo

Turn the research into a list you can act on

These ideas start with finding the right companies. Info finds them and the right person at each, researches every one and drafts the outreach, then asks before it sends.

  • Build this list in Info for idea #14 (opens programmable.info in a new tab)

New accounts: 100,000 free credits with code BETA100K

Find the companies (opens programmable.info in a new tab)
More example playbooks and reviews
ProgrammableDigital

Show Mcpfy’s product working

Turn screenshots or the website into an animated product demo in its own brand, for the homepage, launch posts and outreach in this playbook.

Make a demo (opens programmable.digital in a new tab)

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • FINNY growth strategyAn AI growth platform for financial advisors: it finds and ranks prospects (the F-Score), runs outreach and follow-up, and with the Hunter…22 growth ideas
  • Movers+Shakers growth strategyA social-first creative agency (part of Stagwell): social strategy, campaigns, influencer and creator programs, always-on content and…22 growth ideas
  • Chapman Bright growth strategyA Breda-based B2B consultancy that redesigns marketing and sales operations with automation, AI and agents, specialised in Adobe Marketo…24 growth ideas
  • Sitefire growth strategyA marketing suite for the agentic web: it tracks how AI answers mention a brand, finds the pages and sources that get cited, writes briefs…23 growth ideas
  • Aspire Proptech growth strategyA Bangalore real estate marketing and sales partner: it sells new residential projects for developers and guides buyers through site…21 growth ideas
  • BlackSky growth strategyReal-time space-based intelligence: a Gen-3 satellite constellation, the Spectra tasking and analytics platform, On-Demand and Assured…21 growth ideas
  • MuKn (Mutual Knowledge Systems) growth strategyA Miami startup studio and holding company of technology founders: fractional executives (COO, CFO, CMO, CTO), Salesforce and custom…22 growth ideas
  • CapIntel growth strategyInvestment comparison and proposal software for wealth management firms: advisors compare products and portfolios, show tax impact and…21 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 4, 2026, and cites them throughout. Mcpfy did not commission, review or endorse it, and Programmable is not affiliated with Mcpfy. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how Mcpfy runs its business.

Something wrong, or do you represent Mcpfy? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 307 ideas

    Docs quick start, Trust Center, verifiable proof, store publishing guides, one-click templates and small fixes; the public chat demo.

  2. Days 31 to 609 ideas

    Pulse as a front door, compatibility table, SDK examples, weekly blog, comparisons, upgrade prompts, OpenAPI preview and API outreach; breakage alerts.

  3. Days 61 to 904 ideas

    Commit to the store dashboard, agency partners and the monthly MCP report; run the first build weekend.

The research behind the plan

2 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value git push to a live MCP server. The most cited problem: testing across ChatGPT, Claude and Cursor by hand.

  • Seeing every tool call
  • Getting listed in connector directories
  • In their words: “ship”

Market

Competitive landscape

4 competitors compared: Alpic, Manufact (mcp-use), Smithery and Glama.

  • Gap to exploit: A five-minute path from SDK to store listing
  • Alpic: Head-to-head with Mcpfy on every feature
  • Manufact (mcp-use): Similar positioning

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks firstNew accounts get 100,000 free enrichment credits with code BETA100K (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammableMediaAn AI note taker, interpreter and webinar host for your calls (opens programmable.media in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence