by jonathanmalkin · Astro
Read-only tools for finding where a small business leaks deals, time, and cash.
Read-only tools for finding where a small business leaks deals, time, and cash.
Captured live from the server via tools/list.
START HERE if you are new to Built with Jon. Explains every tool, why to use it, what it returns, how the tools work together, and the best first prompt for finding a leak in deals, time, or cash.
Parameters
No parameters.
{
"type": "object",
"properties": {},
"additionalProperties": false
}List the 15 categories of the Built with Jon Use Case Library (96 worked small-business AI/automation examples), grouped by business function, industry, and personal, with counts and key stats. Call this first to orient before searching.
Parameters
No parameters.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Search the 96-entry Use Case Library of small-business AI and automation workflows. Call this when a user describes a recurring business pain (missed leads, invoice chasing, status meetings, no-shows...) to find worked examples with per-step automation verdicts. Filter by keyword, category, frequency, automation level, or disposition.
| query | string | optional | Keyword matched against name, subject, pain line, workflow steps, and AI-fit text |
| category_id | string | optional | Category id from list_use_case_categories |
| frequency | string | optional | Substring match on frequency, e.g. 'daily', 'weekly', 'monthly', 'per hire' |
| automation | string | optional | |
| disposition | string | optional | Only use cases where some step gets this verdict |
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "Keyword matched against name, subject, pain line, workflow steps, and AI-fit text"
},
"category_id": {
"type": "string",
"enum": [
"admin",
"sales",
"marketing",
"gc",
"cc",
"hs",
"re",
"pm",
"ps",
"hw",
"finance",
"cs",
"hr",
"op",
"personal"
],
"description": "Category id from list_use_case_categories"
},
"frequency": {
"type": "string",
"description": "Substring match on frequency, e.g. 'daily', 'weekly', 'monthly', 'per hire'"
},
"automation": {
"type": "string",
"enum": [
"Fully automatable",
"Mostly automatable",
"Partially automatable"
]
},
"disposition": {
"type": "string",
"enum": [
"eliminate",
"simplify",
"automate",
"optimize",
"report"
],
"description": "Only use cases where some step gets this verdict"
}
},
"additionalProperties": false
}Get one use case from the library in full: the pain, the workflow as it actually runs, per-step verdicts (eliminate / simplify / automate / optimize / report) with rationale, where AI genuinely fits, and the after state. Ids look like 'A1' or 'S3'.
| id | string | required | Use case id, e.g. 'A1' |
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Use case id, e.g. 'A1'"
}
},
"required": [
"id"
],
"additionalProperties": false
}Run Built with Jon's 3-minute AI Workflow Scorecard: ask the user these questions conversationally, then call this with their answers to get a scored verdict on where their business is leaking deals, time, and cash — plus the first fix worth making. Same deterministic scoring as builtwithjon.com/scorecard/. No email or signup involved.
| segment | string | required | What kind of business is this? gc = General contracting; re = Real estate; hs = Home services & trades; pm = Property management; ps = Professional services; hw = Health & wellness; cc = Coaching or creator; general = Something else / general business |
| q2 | string | required | A new lead comes in. How fast does someone respond? |
| q3 | string | required | A lead doesn't buy right away. Then what? |
| q4 | string | optional | Roughly how many new leads or inquiries come in a month? (optional — flavors the reading, does not affect the score) |
| q5 | string | optional | How many hours a week do you spend on admin, chasing status, and re-typing the same information? (optional — flavors the reading, does not affect the score) |
| q6 | string | required | How much of your work lives across separate tools you copy between by hand? |
| q7 | string | required | How often does work get redone because of a miss, a gap, or bad information? |
| q8 | string | required | After work is done, how fast do you invoice and get paid? |
| q9 | string | optional | Last one, and you can skip it. Roughly, what's monthly revenue? (optional — flavors the reading, does not affect the score) |
{
"type": "object",
"properties": {
"segment": {
"type": "string",
"enum": [
"gc",
"re",
"hs",
"pm",
"ps",
"hw",
"cc",
"general"
],
"description": "What kind of business is this? gc = General contracting; re = Real estate; hs = Home services & trades; pm = Property management; ps = Professional services; hw = Health & wellness; cc = Coaching or creator; general = Something else / general business"
},
"q2": {
"type": "string",
"enum": [
"Within minutes",
"Same day",
"A day or two",
"Honestly, it's inconsistent"
],
"description": "A new lead comes in. How fast does someone respond?"
},
"q3": {
"type": "string",
"enum": [
"They go into a follow-up sequence that runs on its own",
"Someone follows up manually, when they remember",
"Usually nothing",
"Not sure"
],
"description": "A lead doesn't buy right away. Then what?"
},
"q4": {
"type": "string",
"enum": [
"Under 10",
"10 to 50",
"50 to 200",
"More than 200"
],
"description": "Roughly how many new leads or inquiries come in a month? (optional — flavors the reading, does not affect the score)"
},
"q5": {
"type": "string",
"enum": [
"A few",
"Around 5 to 10",
"10 to 20",
"More than I want to admit"
],
"description": "How many hours a week do you spend on admin, chasing status, and re-typing the same information? (optional — flavors the reading, does not affect the score)"
},
"q6": {
"type": "string",
"enum": [
"It's mostly connected",
"Some manual copying",
"A lot of manual copying",
"It's all held together by hand"
],
"description": "How much of your work lives across separate tools you copy between by hand?"
},
"q7": {
"type": "string",
"enum": [
"Rarely",
"A few times a month",
"Most weeks",
"Constantly"
],
"description": "How often does work get redone because of a miss, a gap, or bad information?"
},
"q8": {
"type": "string",
"enum": [
"Right away, mostly automated",
"Within a week",
"Slow and manual",
"It slips, and some of it never gets chased"
],
"description": "After work is done, how fast do you invoice and get paid?"
},
"q9": {
"type": "string",
"enum": [
"Under $25k",
"$25k to $100k",
"$100k to $500k",
"More than $500k",
"Skip"
],
"description": "Last one, and you can skip it. Roughly, what's monthly revenue? (optional — flavors the reading, does not affect the score)"
}
},
"required": [
"segment",
"q2",
"q3",
"q6",
"q7",
"q8"
],
"additionalProperties": false
}List the 10 leak calculators (missed calls, slow bids, unbilled change orders, unanswered inquiries, engagement-letter delay, cold DMs, slow quotes, no-shows, first-reply speed, invoice aging) with their input fields and defaults. Use before calculate_leak.
Parameters
No parameters.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Put a monthly/annual dollar figure on one operational leak using Built with Jon's Leak Calculator math (identical to the website's). Call when a user wants to know what missed calls, slow follow-up, no-shows, unbilled change orders, or aging invoices are costing them. Missing inputs fall back to illustrative defaults and are flagged.
| calculator_id | string | required | |
| inputs | object | optional | Numbers keyed by the field names from list_leak_calculators; omit any you don't know |
{
"type": "object",
"properties": {
"calculator_id": {
"type": "string",
"enum": [
"missed-calls",
"slow-bids",
"change-orders",
"inquiries",
"letters",
"cold-dms",
"slow-quotes",
"no-shows",
"first-reply",
"invoice-aging"
]
},
"inputs": {
"type": "object",
"description": "Numbers keyed by the field names from list_leak_calculators; omit any you don't know",
"additionalProperties": {
"type": "number"
}
}
},
"required": [
"calculator_id"
],
"additionalProperties": false
}Get Built with Jon's decision frameworks: the Five Dispositions (the five-question treatment every workflow step gets before anyone automates anything) and the 5-phase implementation process. Call when discussing how to decide what to automate.
| section | string | optional |
{
"type": "object",
"properties": {
"section": {
"type": "string",
"enum": [
"dispositions",
"process",
"all"
]
}
},
"additionalProperties": false
}Get the details of the Hidden Profit Review — Jonathan Malkin's paid, measured review of one recurring business workflow (deliverables, worked sample, waitlist mechanics, capacity). Call when a user wants professional help finding or fixing a profit leak.
Parameters
No parameters.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Search Jonathan Malkin's ~50 articles on small-business workflow leaks, Claude Code infrastructure (the Jules system), AI agents, and AI operations. Filter by keyword and/or tag; returns titles, descriptions, and slugs for get_article.
| query | string | optional | Keyword matched against title, description, tags, and body |
| tag | string | optional | Exact tag match, case-insensitive |
| limit | number | optional | Max results, 1-20 (default 10) |
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "Keyword matched against title, description, tags, and body"
},
"tag": {
"type": "string",
"description": "Exact tag match, case-insensitive"
},
"limit": {
"type": "number",
"description": "Max results, 1-20 (default 10)"
}
},
"additionalProperties": false
}Get one article by slug (from search_articles), including its canonical URL and full markdown text when available.
| slug | string | required | Article slug, e.g. 'missed-call-math-home-services' |
{
"type": "object",
"properties": {
"slug": {
"type": "string",
"description": "Article slug, e.g. 'missed-call-math-home-services'"
}
},
"required": [
"slug"
],
"additionalProperties": false
}Personal site for Jonathan Malkin. AI builder, solo founder.
Built with Astro v6. Hosted on Cloudflare Workers as jonathanmalkin-site. Zero JavaScript by default.
From this repo:
npm run build
npm run preview
Live deploys require both steps:
git push origin main
npm run build
find dist -name .DS_Store -delete
wrangler deploy
Do not pass --assets or --name. wrangler.jsonc owns the Worker and asset
routing, including run_worker_first for the MCP endpoint. CLI asset flags can
replace that configuration and cause static assets to intercept POST /mcp.
In the broader Active-Work workspace, use Scripts/deploy-website.sh to run the push, build, cleanup, and Wrangler deploy sequence.
GitHub push alone does not publish the live site.
Hosted server - connect over the network, no local install.
https://builtwithjon.com/mcpclaude_desktop_config.json
{
"mcpServers": {
"builtwithjon": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://builtwithjon.com/mcp"
]
}
}
}Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.