Marketing helpers tools for AI agents: 16 tools, $0.001 per call
Whizbrew has 16 Marketing helpers tools. One call that returns data costs $0.001; a call that finds nothing or fails costs $0. They are part of one catalogue of 890 tools behind one MCP server and one API key (all groups, pricing).
Example calls
utm.build ($0.001): Builds a campaign link: adds utm_source, utm_medium, utm_campaign (and term, content, id) to a web address by fixed rules, so every link of a team is written the same way. Values are trimmed, lower-cased and spaces become underscores; parameters the address already has are kept, existing utm values are replaced. The answer also says which GA4 default channel the link will be counted in and warns when that is Unassigned, when a value holds personal data or awkward characters. Computed on our server.
curl -s -X POST https://whizbrew.com/v1/call/utm.build \
-H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
-d '{"url":"https://example.com/pricing","source":"Newsletter","medium":"email","campaign":"Spring Sale 2026","content":"header button"}'
utm.parse ($0.001): Reads the tracking out of a link: the UTM values, the click ids ad and mail platforms append (gclid, fbclid, msclkid, ttclid, li_fat_id, mc_cid and others, each named), the GA4 default channel the visit would be counted in, the address without any tracking, and what is wrong (capital letters, spaces, a missing or doubled parameter, values after the #). A link without tracking is free. Computed on our server; the link is not opened (follow a short link with web.redirects first).
curl -s -X POST https://whizbrew.com/v1/call/utm.parse \
-H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
-d '{"url":"https://example.com/pricing?utm_source=newsletter&utm_medium=email&utm_campaign=spring_sale&fbclid=IwAR0abc"}'
headline.score ($0.001): Scores a headline or title from 0 to 100 against a fixed checklist and shows every rule with the points it gave: length, word count, a number, a known format (how-to, list, question, comparison, guide), benefit words, feeling words, the share of common words, whether the subject comes first, and penalties for capitals, exclamation marks and clickbait phrases; with suggestions. A checklist, not a prediction of clicks. The word lists are English. Computed on our server.
curl -s -X POST https://whizbrew.com/v1/call/headline.score \
-H "Authorization: Bearer $WHIZBREW_KEY" -H "content-type: application/json" \
-d '{"text":"7 Proven Ways to Grow Your Newsletter in 2026"}'
Over MCP the same call is call_tool with {"tool": "<name>", "arguments": {…}}.
Every tool in this group
| Tool | What it returns | Price per call that returns data | Per 1,000 calls |
|---|---|---|---|
utm.build | Builds a campaign link: adds utm_source, utm_medium, utm_campaign (and term, content, id) to a web address by fixed rules, so every link of a team is written the same way. Values are trimmed, lower-cased and spaces become underscores; parameters the address already has are kept, existing utm values are replaced. The answer also says which GA4 default channel the link will be counted in and warns when that is Unassigned, when a value holds personal data or awkward characters. Computed on our server. | $0.001 | $1 |
utm.parse | Reads the tracking out of a link: the UTM values, the click ids ad and mail platforms append (gclid, fbclid, msclkid, ttclid, li_fat_id, mc_cid and others, each named), the GA4 default channel the visit would be counted in, the address without any tracking, and what is wrong (capital letters, spaces, a missing or doubled parameter, values after the #). A link without tracking is free. Computed on our server; the link is not opened (follow a short link with web.redirects first). | $0.001 | $1 |
headline.score | Scores a headline or title from 0 to 100 against a fixed checklist and shows every rule with the points it gave: length, word count, a number, a known format (how-to, list, question, comparison, guide), benefit words, feeling words, the share of common words, whether the subject comes first, and penalties for capitals, exclamation marks and clickbait phrases; with suggestions. A checklist, not a prediction of clicks. The word lists are English. Computed on our server. | $0.001 | $1 |
subject_line.check | Checks an email subject line (and its preheader) before sending: length and where phone and desktop inboxes cut it, capitals, punctuation, phrases tied to bulk promotions, a fake Re: or Fw:, emoji, unfilled merge tags, whether the point comes first, and the preheader's length. Each check says pass, warn or fail with what it found; the score is 100 minus 10 per warning and 25 per fail. It reads the words only: inbox placement depends far more on the sender (see email.domain_health). Computed on our server. | $0.001 | $1 |
social.text_check | Does this text fit? One post, caption, title or ad text against the length limits of X, Instagram, Facebook, LinkedIn, TikTok, Threads, Bluesky, Mastodon, YouTube, Pinterest, Reddit, Snapchat, Google Ads, Meta ads, search results (title, meta description), an email subject, a push notification and SMS (with the number of message parts). Each row counts the way its platform does (X counts a link as 23 and an emoji as 2), says how many characters are left or over, and where the feed cuts the text behind "more". Also counts hashtags, mentions, links and emoji. Computed on our server from the limits the platforms publish. | $0.001 | $1 |
qr.make | A QR code for a link or a text, as an SVG image (the SVG text and a data: URI ready for an <img> tag), with the size, colours and error correction you choose. The code holds the data itself: it never expires and no scan is counted by anybody; put a link from utm.build in it to see scans in analytics. Made on our server, nothing is sent anywhere. | $0.001 | $1 |
content.calendar | The dates a content or campaign plan is built around, for one country and a date range (default: the next 90 days): the country's public holidays with their local names and regions, the United Nations' international days (World Health Day, International Women's Day...), and the retail dates computed by rule (Black Friday, Cyber Monday, Mother's and Father's Day by country, Valentine's Day, Singles' Day...). Each row has the date, the weekday, how many days away it is, its kind and where it comes from. Public holidays cover about 120 countries; a range with no date is free. | $0.001 | $1 |
abtest.significance | Did variant B really beat A? The two-proportion z-test on the visitors and conversions of two variants: each conversion rate, the absolute and relative difference, z, the two-sided p-value, whether the difference is significant at the confidence you ask for (default 95%), the interval of the difference and the winner (or none). Flags counts too small for the test. Computed from your numbers; nothing is read. | $0.001 | $1 |
abtest.sample_size | How many visitors an A/B test needs before it starts: from the baseline conversion rate and the smallest change worth detecting (relative by default: 20 = a 20% lift; or absolute percentage points), at a confidence (default 95%) and power (default 80%). Returns visitors per variant and in total, and with daily_visitors the days and weeks the test must run. Computed; nothing is read. | $0.001 | $1 |
ads.metrics | The ratios of a campaign from its raw numbers: give any two or more of spend, impressions, clicks, conversions and revenue and get every metric they allow: CPM, CTR, CPC, conversion rate, CPA, ROAS, average order value, revenue per click and per 1000 impressions, and revenue minus spend. A metric whose inputs are missing is null and named in not_computed. Computed; nothing is read. | $0.001 | $1 |
ads.break_even | What an ad may cost before it loses money: from the order value and the margin (or the unit cost), the highest cost per acquisition, the lowest ROAS and, with a conversion rate, the highest cost per click, for the first order and for the customer's lifetime (orders_per_customer). With target_profit_pct the same limits at the profit you want to keep. Computed; nothing is read. | $0.001 | $1 |
saas.unit_economics | The unit economics of a subscription from four numbers: customer lifetime (1 / churn), lifetime value with gross margin, the LTV to CAC ratio, the months to earn the acquisition cost back, the yearly churn that the monthly churn means, and the highest CAC a 3:1 ratio allows. Says how the result reads against the usual rules of thumb. Assumes constant churn and no expansion revenue. Computed; nothing is read. | $0.001 | $1 |
survey.sample_size | How many answers a survey needs: for a confidence (default 95%) and a margin of error (default 5 points), with the correction for a small population (a customer list of 2,000) and, from the response rate you expect, how many people to invite. Give responses to get the margin of error of the answers you already have. Computed; nothing is read. | $0.001 | $1 |
survey.nps | The Net Promoter Score from the counts of promoters (9-10), passives (7-8) and detractors (0-6): the score from -100 to 100, each share, and the 95% margin of error with its interval, so two scores can be compared honestly. Flags a sample under 30. Computed; nothing is read. | $0.001 | $1 |
color.contrast | Is this text readable on this background? The WCAG 2 contrast ratio of two colours and what it passes: AA and AAA for normal and large text, AA for buttons and icons. When it fails, the nearest darker or lighter shade of the text colour that passes. For ad creatives, landing pages, emails and brand palettes. Computed; nothing is read. | $0.001 | $1 |
brand.icon | A brand's single-colour mark and main brand colour by brand name or website, from the Simple Icons set of about 3,400 brands: name, the colour as a hex code, the address of the mark as an SVG (in the brand colour and in black), the brand's own page the mark was taken from, its usage guidelines and licence when recorded. An exact name first; a start of a name lists the brands that begin with it. For slides, comparison pages and partner logos. A brand that is not in the set answers with no matches and is free; web.brand_assets reads any site's own logo, colours and fonts. One data file read at most once a day, no login, no key. | $0.001 | $1 |
Inputs, output shape, health and measured success rate of each tool: https://whizbrew.com/v1/tools and https://whizbrew.com/status#marketing.
Start
- An agent on its own: read https://whizbrew.com/agents.md. It asks for a free quote (
GET https://whizbrew.com/v1/agent/guide?goal=<the task in plain words>), then one request (POST https://whizbrew.com/v1/agent/start) gives it a key that works at once with the free credit, with no email and no sign-up form. Its human claims the account later. - A chat app (ChatGPT, Claude, Gemini): add the MCP server
https://whizbrew.com/mcpas a custom connector and sign in. No key to copy. - An agent with a shell (Claude Code, Codex, Cursor): read https://whizbrew.com/llms.txt. The agent starts a pairing with one request, its user approves with one click, and the agent holds its own key.
- Plain HTTP: create an account at https://whizbrew.com/app#signup, make a key, then
POST https://whizbrew.com/v1/call/<tool>withAuthorization: Bearer <key>.