# FindUtils MCP [Health: Active] **Category:** πŸ’» Developer Tools **Repository:** https://mcp.findutils.com/ **Views:** 1 **Installs:** 0 **Upvotes:** 0 **Directory Page:** https://allmcps.com/mcp/findutils-mcp ## Description Remote MCP server exposing 388 FindUtils tools (JSON/YAML/CSV, encoding, hashing, converters, generators) to AI agents. ## Tools Capabilities this server exposes over MCP: - **invoice_generate** β€” Generate a professional PDF invoice. Returns a base64-encoded PDF. Supports line items, tax, discounts, notes, and full sender/recipient details. - **a2ui_spec** β€” Returns A2UI protocol documentation for a given spec version and section. Sections: protocol, components, data-binding, actions, catalogs, custom-functions, renderers, renderer-contract, transports, mcp-apps, agents, sdk-architecture, ecosystem. Use this when authoring A2UI payloads to get authoritative reference material without guessing. - **a2ui_basic_catalog** β€” Returns the authoritative basic-catalog JSON Schema (components, functions, theme) for the requested A2UI version. Use this to validate agent output or bootstrap a custom catalog. - **a2ui_examples** β€” Returns one or more canonical A2UI JSONL examples matching a pattern (form, list, wizard, modal, chat, dashboard, checkout, typeahead, mcp-card, multi-surface, theme-*). Use these as few-shot examples when prompting an LLM to generate A2UI. - **a2ui_diff** β€” Computes a structural diff between two A2UI JSONL payloads: surfaces added/removed, components added/removed/changed (by id), data-model path-level changes. Use this to understand what an agent turn changed in a surface, or to verify that a refactor kept behaviour identical. - **a2ui_prompt** β€” Generates a drop-in LLM system prompt that teaches the model to output valid A2UI JSONL: includes the basic-catalog component / function list, the hard rules (one root, IDs not inline, paths start with "/"), and the requested few-shot examples. Feed the returned string straight into your chat-completion call's system role. - **a2ui_validate** β€” Validates a JSONL A2UI payload against the basic catalog (or a custom catalog) and returns any errors, warnings, or info messages with exact line numbers. Call this on every LLM-generated payload before sending to the client. - **pdf_extract** β€” PDF Text Extractor. Extract text and structured JSON (with bounding boxes) from a PDF. Accepts either a public https:// URL pointing to a PDF, or a base64-encoded PDF string. Base64 always works; some sites refuse requests that come from a server (HTTP 403), so send those files as base64. The PDF is parsed with WebAssembly and its bytes are not logged. Best for documents under 10MB. - **age_calculate** β€” Calculate age from a date of birth β€” years, months, days, plus totals in days, weeks, and months. Accepts ISO 8601 or anything Date.parse handles. - **ai_model_picker** β€” Return the best-matching AI coding models (top 3 by default) for a project, scored on speed, accuracy, cost and context window from four answers: what you build, primary language, top priority and project complexity. Model data comes from the live FindUtils model catalog. - **amortization_calculator** β€” Return a full loan amortization schedule (one row per month with payment, principal, interest, extra payment, balance) plus totals, interest saved, months saved, and the payoff month. Rate is a yearly percentage; payments are monthly. - **api_docs_generator** β€” Return API documentation generated from a list of endpoints, as Markdown (with table of contents, request/response examples) or as an OpenAPI 3.0.3 JSON specification. - **aspect_ratio** β€” Compute the aspect ratio of width:height. Optionally scale to a new width or height. - **ass_to_srt** β€” Flatten an Advanced SubStation Alpha (.ass) or SubStation Alpha (.ssa) subtitle script to plain SubRip (.srt). Reads the Dialogue lines of the [Events] section using the Format line for column order, converts H:MM:SS.cc centisecond timestamps to HH:MM:SS,mmm, strips every {\...} override block, turns \N into a line break, sorts cues by start time and renumbers them from 1. Styles, fonts, colours, positioning and drawing commands are discarded on purpose, because SubRip cannot carry them. Comment lines are ignored. - **attendance_sheet_generator** β€” Generate a blank attendance sheet for a list of students over consecutive days and return the date columns, a CSV, a Markdown table, and a plain-text grid. Dates are consecutive calendar days from start_date (UTC), labelled like "Mon, Mar 2". - **auto_loan_calculator** β€” Return the monthly car payment, total interest, financed amount, and a month-by-month amortization schedule for an auto loan. Accounts for down payment, trade-in value, and sales tax (optionally financed). Rate is a yearly percentage. - **avatar_generator** β€” Make an avatar as an SVG from a name or an email address, with no external service. Initials mode draws up to two initials on a background colour derived from the input; identicon mode draws a mirrored 5Γ—5 block pattern. Both are seeded from a SHA-256 of the trimmed, lower-cased input, so the same input always gives the same avatar. Returns the SVG, a data: URL, the colours and the initials or the grid. - **barcode_generator** β€” Generate a 1D barcode as an SVG string from text. CODE128 takes any printable ASCII, CODE39 takes A-Z, 0-9, space and -.$/+%. EAN13, EAN8 and UPC (UPC-A) take digits: give 12, 7 or 11 digits and the GS1 check digit is appended, or the full code and it is verified. Returns the SVG, the bar/space bit pattern, the printed text, and the image size. - **base32_decode** β€” Decode RFC 4648 base32 back to UTF-8 text. Case-insensitive; padding "=" optional. - **base32_encode** β€” Encode UTF-8 text to RFC 4648 base32 (uppercase, "=" padding). Useful for TOTP secrets and case-insensitive identifiers. - **base64_decode** β€” Decode a base64 string back to UTF-8 text. - **base64_encode** β€” Encode UTF-8 text as a base64 string. - **bmi_calculate** β€” Compute Body Mass Index (BMI) with metric (kg/cm) or imperial (lb/in) units. - **body_fat_calculator** β€” Estimate body fat percentage, fat mass, lean mass, BMI, and a body-composition category. Four methods: the U.S. Navy circumference method (waist, neck, hip), a BMI-based estimate (Deurenberg), and the Jackson-Pollock 3-site and 7-site skinfold equations with Siri conversion. Metric (kg/cm) or imperial (lbs/in) inputs; skinfolds are always millimetres. - **bookmarks_diff** β€” Compare two Netscape bookmarks.html exports (before and after) and list the links that were added, removed, or changed. Links are matched by address the way dedupe_bookmarks compares them: host case, default port and trailing slash folded, and by default utm_* / fbclid / gclid parameters and the #fragment ignored. A match is "changed" when its title differs or, with compare_folders (default true), when it sits in a different folder. Returns counts, one entry per difference, a CSV report, and the added and removed links as importable bookmarks.html files. An address saved twice on one side is compared once, by its first copy. No link is fetched. - **bookmarks_html_to_csv** β€” Convert a browser bookmarks.html export from Chrome, Firefox, Edge, Safari or any other browser that writes the Netscape bookmark format into CSV, one row per link. Each row carries the folder path flattened with " / ", the title, the URL, and the ADD_DATE and LAST_MODIFIED stamps as ISO-8601 when the export has them. Favicon data stored in the ICON attribute is never written to the CSV and no favicon is ever fetched. Nothing is checked against the network, so a dead link is reported exactly as the export stored it. The browser page accepts a 10 MB file; the API and MCP surfaces cap a request at 1 MB, so a very large export belongs on the page. - **bookmarks_html_to_markdown** β€” Convert a browser bookmarks.html export (the Netscape bookmark format Chrome, Firefox, Edge and Safari write) into Markdown for Notion, Obsidian or a README. In the default "headings" style each folder becomes an ATX heading nested by depth (## for a top-level folder, ### below it, up to ######) and each bookmark a "- [title](url)" line under its folder, in the order the export lists them; "list" style writes one flat bullet list with the folder path before each link. The document starts with a # heading named after the export (or the title argument), and front_matter adds a YAML block with that title. A bookmark with no title is named after its host. Entities are decoded, Markdown characters in titles are escaped, favicon data is never written, and no link is fetched. An ordinary web page is refused: it has links but no bookmark folder structure. - **bookmarks_json_to_html** β€” Convert a bookmarks JSON file into a Netscape bookmarks.html that Chrome, Firefox, Edge and Safari import. Reads three shapes: the Chromium profile file named Bookmarks (Chrome, Edge, Brave; no extension; roots.bookmark_bar / other / synced, date_added in microseconds since 1601), a Firefox JSON backup (text/x-moz-place containers with uri and dateAdded in microseconds since 1970), and a plain tree of {title|name, url|uri|href, children} objects or an array of them. Folder nesting and root folders are kept, dates are converted to the Unix seconds bookmarks.html uses, entries with no address and separators are skipped and counted. Other formats are refused, and nothing is fetched. - **bot_crawl_checker** β€” Return which of 24 well-known crawlers (Googlebot, BingBot, GPTBot, ClaudeBot, PerplexityBot, AhrefsBot, Twitterbot and more) are allowed or blocked for a URL by the site's robots.txt, with the matching rule and user-agent group per bot. - **box_shadow_generator** β€” Generate a CSS box-shadow declaration from offset, blur, spread, colour, opacity, and inset settings, or from a named preset (subtle, soft, medium, large, xl, inner, glow, hard). Returns the full "box-shadow: ...;" rule, the bare value, the rgba colour, and the resolved settings. - **break_even_calculator** β€” Return the break-even point of a product or business: contribution margin, contribution margin ratio, break-even units and revenue, plus the units and revenue needed to reach a target profit. Units are rounded up to whole units. - **business_loan_calculator** β€” Return the monthly payment, total interest, origination fee, effective yearly rate (interest plus fees), a yearly summary, and a full monthly amortization schedule for a business loan. Rate is a yearly percentage; the term is in months. - **calorie_calculator** β€” Work out how many calories to eat per day to lose, maintain, or gain weight. Maintenance comes from the Mifflin-St Jeor BMR and an activity multiplier (the same chain as the TDEE calculator), then a mild/moderate/extreme rate applies a 250, 500, or 1000 kcal daily deficit or surplus. Also returns the projected weekly weight change and a protein/carbs/fat split. - **caption_generator** β€” Generate a social-media caption from a topic using built-in templates (no AI): picks a template for the content type and tone, fills in the topic, and optionally appends a call to action and up to 8 hashtags. Returns the caption, its length, and the platform limit. - **carbon_footprint_calculator** β€” Estimate an annual personal carbon footprint in tonnes of CO2 from home energy, transportation, and lifestyle, with a per-category breakdown and a comparison to the US average (16 t), the world average (4.7 t), and the Paris target (2 t). Every input is optional and defaults to a typical US household value. - **career_experience_calculator** β€” Calculate total professional experience across several jobs: total days, years/months/days, weeks, working days (5/7), working hours (8 per working day), per-job durations, and career milestones (1 to 25 years). Jobs with an invalid or reversed date range are skipped and counted. - **case_convert** β€” Convert a string between camelCase, PascalCase, snake_case, kebab-case, SCREAMING_SNAKE, dot.case, Title Case, Sentence case, lower, UPPER. - **chmod_calculate** β€” Convert between numeric (755) and symbolic (rwxr-xr-x) Unix file permissions. - **cidr_calculate** β€” Compute network / broadcast / first-last host / subnet mask / hosts count for a CIDR (e.g. 192.168.1.0/24). - **citation_generator** β€” Generate a reference-list citation in APA 7, MLA 9, Chicago, or Harvard style for a book, website, journal, or article. Returns the citation with *italic* markers, a plain-text version, and an HTML version with tags. - **client_status_report_generator** β€” Generate a client-facing project status report in Markdown and return it with its headline counts. Sections: header (client, period, report type, generated date), timeline status, budget spent vs total, an executive summary, completed work, in-progress items with percent complete, active blockers with severity, upcoming goals, and notes. Empty sections are omitted, exactly as the page's export does. - **clip_path_generator** β€” Generate a CSS clip-path from a named shape (triangle, trapezoid, parallelogram, rhombus, pentagon, hexagon, octagon, star, chevron, message, circle, ellipse, inset) or from your own percentage coordinate pairs. Coordinates are relative to the element box, so the shape scales with the element. Returns the full "clip-path: ...;" rule, the bare value, and the resolved points. Note that clip-path cuts the element visually but does not change its layout box. - **cloudflare_cost_calculate** β€” Estimate monthly Cloudflare developer platform costs (Workers, D1, R2, KV, Durable Objects) under free or $5 Workers Paid plans. - **code_diff_checker** β€” Returns a line-by-line diff of two code snippets: each line marked added, removed or unchanged with its original and modified line numbers, plus addition/deletion counts and a unified diff (diff -u format, with @@ hunk headers and 3 lines of context) that patch and git apply accept. Can ignore whitespace and letter case when comparing. - **code_highlight_html** β€” Returns syntax-highlighted HTML for a code snippet: HTML-escaped text wrapped in elements using one of six colour themes (dracula, monokai, github, nord, one-dark, solarized). This is the pure highlighting step of the Code Screenshot Generator page; it returns markup plus the theme background and foreground colours, not an image. - **color_contrast** β€” Compute the WCAG 2.1 contrast ratio between two colours (accepts #rgb, #rrggbb, and rgb()/rgba() formats) and report whether it passes AA/AAA for normal and large text. - **color_convert** β€” Convert between color formats: hex, rgb, hsl, hsv. - **color_palette_generator** β€” Generate a harmonious colour palette from a base hex colour: complementary, analogous, triadic, tetradic, split-complementary, or monochromatic. Returns the hex list, per-colour rgb/hsl swatches with a readable text colour, and a CSS custom-properties block. - **compound_interest** β€” Compute compound interest growth. Supports any compound frequency (monthly, quarterly, annual) plus optional fixed monthly contributions. Returns final balance, total contributions, total interest, and a year-by-year balance series. - **cookie_analyzer** β€” Analyze Set-Cookie header strings and return a per-cookie list of security issues, an overall 0-100 score, and recommendations. Checks Secure, HttpOnly, SameSite, expiry, path scope, sensitive names, and the __Secure-/__Host- prefix rules. One cookie per line; the "Set-Cookie:" prefix is optional. - **country_lookup** β€” Return country facts β€” capital, currency (name, code, symbol), phone code, timezone, region, subregion, flag emoji, and languages β€” for countries that match a search query, an ISO 3166-1 alpha-2 code, or a region. Call with no filters to list every country. - **coupon_code_generator** β€” Generate a batch of unique random coupon or promo codes: an optional prefix, a random part of a chosen length and character set (with an option to drop the look-alikes 0, O, 1, I and L), and optional groups split by a separator. Codes come from crypto.getRandomValues with unbiased sampling, every code in a batch is different, and a request for more codes than the pattern can make is refused. Up to 10,000 codes per call. - **course_quiz_generator** β€” Build a printable course quiz from a list of questions (multiple choice, true/false, short answer, fill in the blank) and return the numbered questions with labels, the total points, an answer key, a plain-text version, and a standalone HTML document. LaTeX between $…$ is kept as-is (not rendered). - **credit_card_payoff_calculator** β€” Return how long a credit card balance takes to pay off at a fixed monthly payment (mode "payment"), or the payment needed to clear it in a target number of months (mode "months"), with total interest, total paid, payoff month, and a monthly schedule. APR is a yearly percentage. - **credit_card_validate** β€” Validate a credit card number with the Luhn checksum and detect the brand. - **cron_describe** β€” Parse a 5-field cron expression (minute hour day-of-month month day-of-week) and produce a human-readable English description. - **cron_to_systemd** β€” Convert a 5-field cron expression into a systemd timer plus service unit and a Kubernetes CronJob YAML file. Six-field cron (with seconds) is rejected. - **crypto_testnet_faucets** β€” Returns a curated list of crypto testnet faucets (Ethereum Sepolia, Base, Arbitrum, Optimism, Polygon, Avalanche, BNB Chain, Linea, Solana) with the faucet URL, tokens dispensed, amount, cooldown, auth requirements and USDC contract where known, plus the chain config (chain id, RPC URL, explorer) for each listed network. Filter by network and a search term. - **csp_generate** β€” Build a Content-Security-Policy header from a preset (strict, typical, google-analytics) and/or your own directives, and get the header line, an equivalent tag, and warnings such as script-src allowing 'unsafe-inline', 'unsafe-eval', * or data:, a missing object-src 'none', base-uri or frame-ancestors. Bare keywords like self are quoted for you; invalid sources are rejected with the directive named. Supports Report-Only with a report-uri. Pure computation: it checks the policy text, not a live site. - **css_animation_generator** β€” Return CSS @keyframes animation code (or Tailwind config, Web Animations API / React, or Framer Motion code) from a named preset such as fade-in, bounce, pulse, shake, spin, or from custom keyframes with transform, opacity, filter and color properties. - **css_filter_generator** β€” Generate a CSS filter declaration from brightness, contrast, saturate, grayscale, sepia, invert, opacity, blur and hue-rotate, or from a named preset (vintage, cool, warm, dramatic, grayscale, faded, sharpen). Only functions that differ from their no-op value are emitted, so the output stays short. Returns the full "filter: ...;" rule, the bare value, and which functions were applied. Order matters in CSS filters and is preserved as listed. - **css_grid_generator** β€” Generate a CSS Grid container declaration. sizing=equal gives repeat(n, 1fr); auto-fit and auto-fill give a responsive repeat(auto-fit, minmax(min_width, 1fr)) that needs no media query; custom takes your own track lists. Optional named areas are validated so every row declares the same number of cells, which is the usual reason a grid silently collapses. Returns the full rule, the separate template values, and the area names for the child rules. - **css_minifier** β€” Minify CSS (strip comments and whitespace) or beautify compressed CSS, and return the output with byte sizes and the percentage saved. Minify removes comments, whitespace around punctuation, and the last semicolon in each block. Beautify puts one declaration per line with the chosen indentation. - **csv_dedupe** β€” Remove duplicate rows from a CSV or TSV table. With no keys, a row is a duplicate when every cell matches an earlier row; with keys (column names), when those columns match, so rows with the same email but a different name count as duplicates. keep "first" (default) keeps the earliest occurrence and "last" keeps the latest, and either way the kept rows stay in their original file order. trim (default true) ignores surrounding spaces when matching and ignore_case (default false) ignores letter case; cells are written out unchanged. A row whose key cells are all blank is kept as unique unless blank_keys is "collapse". Returns cleaned_csv (header + kept rows) and duplicates_csv (header + removed rows) in the input delimiter, counts, and a preview of removed rows. Quoted fields, embedded commas and line breaks are preserved; a quote that never closes stops with UNCLOSED_QUOTE and the line it opens on. Nothing is uploaded when used on the browser page. - **csv_diff** β€” Compare two CSV or TSV tables by key column and list the rows that were added, removed or changed - a table diff, not a line diff, so rows in a different order are not changes. keys names one or more columns (a composite key, e.g. email + country); when omitted the tool picks a shared column named id, else a unique *_id column (customer_id), else one named email, sku, uuid, key or code, else the first shared column whose values are unique in both files, else it falls back to whole-row comparison and says so in notes. A row is changed when a non-key column present in both files differs (cells compared exactly; trim and ignore_case apply to key matching only). Duplicate keys within a file keep the first row and are reported in duplicate_keys. A quote that opens and never closes stops the comparison with UNCLOSED_QUOTE naming the file and line. mode "row" compares whole rows as a multiset, with no changed class. Returns counts, the rows (capped at max_rows), and diff_csv: comma-separated, header change_type, then every column of A followed by the new columns of B, then old_ for each column that changed anywhere; removed rows hold A's values, added and changed rows B's, and old_ holds A's value on changed rows. diff_csv lists every added, removed and changed row; unchanged rows are left out. Nothing is uploaded when used on the browser page. - **csv_to_bookmarks_html** β€” Convert a CSV of links into a Netscape bookmarks.html file that Chrome, Firefox, Edge and Safari import. Columns are matched by name: url or href or link, title or name, folder or folder_path or path, and an optional add_date, so a sheet exported from a bookmark manager usually needs no configuration; column_map overrides any of them. A folder path builds nested folders, split on " / " when the value uses it and on "/" otherwise, so a folder genuinely called "AI/ML" survives. Every row needs an address; a row without one is skipped and listed by row number. A bare domain gains https://. Nothing is fetched and no link is checked. - **csv_to_ics** β€” Turn a CSV of events into an iCalendar (.ics) file that Google Calendar, Apple Calendar and Outlook import. Columns are matched by name: Subject or Title, Start Date, Start Time, End Date, End Time, All Day Event, Description, Location and a few aliases, so a Google Calendar or Outlook export maps with no configuration; column_map overrides any of them. Each row needs a title and a start. A row with no time, or All Day Event set, becomes an all-day event. Times are written as stated; an optional IANA time zone is stamped on them, UTC writes Z, and no zone writes floating time. Rows without a title or a readable start are skipped and listed. Recurrence is not inferred from text. - **csv_to_internet_shortcuts** β€” Turn a CSV of names and addresses into Windows Internet Shortcut (.url) files, one per row. Returns each file as { name, url, content }, where content is the .url text ([InternetShortcut] then URL=, CRLF line endings), so a caller can write them to disk or zip them. The url column is found by its header (url, link, href, address, website) and the name column by name, title or label; url_column and name_column override the match. A CSV with no header row is read when a cell of its first row is an address. A bare domain gains https://. A row with no address, an address that is not one, or (with web_only, the default) a non-http(s) address such as file: is skipped with its row number and reason, and the other rows still convert. A row with no name is named after its host. Names are made safe for Windows and macOS (\ / : * ? " < > | removed, reserved names such as CON suffixed), and a repeated name gets " (2)". Addresses are written as given and never opened. Does not create .lnk files. - **csv_to_json** β€” Parse CSV text into an array of objects keyed by the header row. - **csv_to_ndjson** β€” Convert CSV into newline-delimited JSON (NDJSON, JSON Lines, .jsonl): one JSON object per line, keyed by the header row. Quoted fields, embedded line breaks and doubled quotes are read per RFC 4180; the delimiter is detected from the header when not given. Every value stays a string unless coerce is true, which turns true, false, null, an empty cell and plain numbers into JSON values while leaving leading zeros and integers of 2^53 or more as text. An empty header name becomes column_N and a duplicate gets a numeric suffix. A row wider than the header keeps the first N values and is reported by row number. Dotted header names are not rebuilt into nested objects. - **csv_to_sql** β€” Convert CSV text into SQL INSERT statements, optionally preceded by a CREATE TABLE. The first row is the header and becomes the column list. Column types are inferred from the first 50 data rows as integer, numeric, boolean, or text, and identifiers are quoted for the chosen dialect (Postgres, MySQL, or SQLite). Empty cells become NULL by default. Nothing is executed against a database. - **csv_to_tsv** β€” Convert comma-separated values to tab-separated values, the format spreadsheet paste, psql and mysql --batch expect. Quoted fields, embedded line breaks and doubled quotes are read per RFC 4180; the input delimiter is detected from the first line when not given. A cell that holds a tab, a quote or a line break is quoted the way Excel's tab-delimited export quotes it; every other cell is written bare. Values stay text. A leading byte order mark is dropped and blank lines are skipped. - **csv_to_vcf** β€” Convert a CSV of contacts into one multi-card .vcf file that iPhone, Android, Google Contacts and Outlook import. Columns are matched by name: full name or first and last name, email, phone or mobile, work and home phone, company, job title, website, notes, street, city, region, postal code, country and birthday, including the headers Google Contacts, Outlook and vcf_to_csv write; column_map overrides any match. One card per row with a name, email or phone; other rows are skipped and listed by row number. vCard 3.0 by default, 4.0 on request. Values are escaped and lines folded at 75 octets per RFC 6350. Photos are not written and nothing is uploaded or synced. - **curl_to_code** β€” Returns source code that performs the same HTTP request as a cURL command, in JavaScript (fetch), Python (requests), PHP (cURL), Go (net/http), Ruby (net/http) or Node.js (https). Also returns the parsed request: method, URL, headers, body and basic auth. Parses -X, -H, -d/--data/--data-raw/--data-binary, -F/--form (name=value only; file fields are rejected), -u and the URL. - **daily_habit_calculator** β€” Calculate how a daily habit adds up over a week, a month (30 days), a year (365 days), or a custom date range: total units, total cost, and per-week/month/year amounts. Works for coffee cups, water glasses, steps, pages, cigarettes, or any custom unit. - **data_sanitizer** β€” Sanitize untrusted text and return the cleaned output, a list of changes made, and a risk level. Modes: html (escape HTML special characters), sql (escape quotes/backslashes, strip comments and semicolons), js (strip script tags, javascript: URLs, event handlers, CSS expressions), url (percent-encode), or all (html + sql + js). Defaults to all. - **date_difference** β€” Compute the difference between two dates in years/months/weeks/days/hours/minutes/seconds. - **debt_payoff_calculator** β€” Return a debt payoff plan for several debts using the avalanche (highest rate first) or snowball (lowest balance first) strategy: total months, total interest, total paid, and the order and month each debt is cleared. Minimum payments roll into the extra payment as debts close. - **dedupe_bookmarks** β€” Remove duplicate links from a Netscape bookmarks.html export (Chrome, Firefox, Edge, Safari) and return a clean bookmarks.html plus a CSV of the duplicates. Two bookmarks are duplicates when their addresses match after folding letter case in the host, default ports and a trailing slash, and, by default, after dropping utm_* / fbclid / gclid tracking parameters and the #fragment. The first copy is kept with its title, date and folder; the duplicates CSV lists every folder a copy sat in. Other query parameters are never touched, non-web addresses compare verbatim, and no link is fetched. - **dev_request_prioritizer** β€” Prioritize development requests with the page's scoring frameworks and return them sorted by priority score, each with a value score and an effort/value quadrant (quick win, big bet, fill-in, money pit), plus a top recommendation and totals. Frameworks: "default" (weighted value Γ· √effort over Customer Impact, Revenue Impact, Strategic Value, Urgency, Risk), "rice", "ice", "wsjf", "moscow". Score each request 1-10 per dimension, keyed by dimension name; optional confidence weighting and a deadline boost (+20% within 7 days, +50% within 3, +100% overdue, capped by deadline_boost_multiplier). - **dividend_calculator** β€” Return a year-by-year projection of a dividend portfolio: dividends received, DRIP reinvestment, contributions, share count, share price, ending value, plus final yearly and monthly dividend income and yield on cost. All rates are yearly percentages; contributions and reinvested dividends buy shares at the mid-year average price. - **dnpm_configurator** β€” Returns the five files of a hardened Docker-based npm wrapper (the ./dnpm bash script, .dnpm/Dockerfile, docker-compose.node.yml, .dnpm/seccomp-profile.json and a CLAUDE.md) generated for a project. The sandbox mounts the source read-only, drops all capabilities, runs as a non-root user, disables postinstall scripts, pins the npm registry and builds with zero network access. - **dns_lookup** β€” Return the DNS records of a domain for one record type (A, AAAA, CNAME, MX, TXT, NS or SOA). Resolved through the public Google DNS-over-HTTPS resolver. - **dns_propagation_checker** β€” Compare one DNS record type for a domain across two named public DNS-over-HTTPS resolvers (Google Public DNS at dns.google and Cloudflare at cloudflare-dns.com): each resolver's response code, values and TTLs, whether they agree, and which resolver returned each value. Two resolvers are not a global propagation map. - **dns_security_scanner** β€” Return a DNS security report for a domain: SPF, DMARC, MX and CAA records with issues, a 0-100 score, a letter grade and recommendations. Records are resolved through Cloudflare DNS-over-HTTPS. - **domain_rating_checker** β€” Return the Ahrefs Domain Rating (0-100) for up to 25 domains in one call, with a per-domain status (provider_unavailable means Ahrefs is not answering; when every domain is in that state the call fails with PROVIDER_UNAVAILABLE). Results carry the required "Domain Rating by Ahrefs" attribution. - **down_payment_calculator** β€” Return the down payment, closing costs, total cash needed, remaining amount to save, months to save, target month, loan amount, and estimated PMI for a home purchase, plus a comparison table for 5%, 10%, 15%, 20%, and 25% down. PMI is estimated at 1% of the loan per year when the down payment is below 20%. - **eisenhower_matrix** β€” Sort tasks into the four Eisenhower quadrants and say what to do with each one. Every task is urgent or not and important or not: urgent + important is "Do First", important only is "Schedule", urgent only is "Delegate", and neither is "Eliminate". Pass `urgent` and `important` per task, or give a `due` date and let urgency follow from how close it is. Returns the four quadrants with their tasks, per-quadrant counts, and the single task to start on. - **email_header_analyzer** β€” Return a structured analysis of raw email headers: sender, recipient, subject, date, message id, the Received hop chain with per-hop delays, SPF/DKIM/DMARC results, every public IP seen, and suspicious-routing flags. - **email_security_checker** β€” Return an email-security report for the domain of an email address: MX, SPF, DKIM (common selectors), DMARC, MTA-STS, TLS-RPT and BIMI checks with a 0-100 score and recommendations. Records are resolved through Cloudflare DNS-over-HTTPS. - **email_signature_generator** β€” Generate an HTML email signature (table-based, inline styles, safe for Gmail/Outlook) from name, title, company, contact details and social links. Returns the HTML in a horizontal, vertical, or compact layout. - **email_validate** β€” Validate an email address and report every rule it breaks and break it into local / domain / TLD parts. - **env_linter** β€” Lint a dotenv file for duplicate keys, unquoted spaces, export prefixes, secret-looking values (masked), and drift against an optional .env.example. Returns issues and a masked key table. Does not execute the file. - **env_yaml_converter** β€” Convert a dotenv .env file to a flat YAML map, or YAML back to KEY=value lines. The shape option wraps the map as a pasteable Docker Compose (services.app.environment) or GitHub Actions (jobs.app.env) snippet rather than a whole file. Comments, blank lines and an "export " prefix are handled, and a duplicate key keeps the last value and is reported. Values stay strings unless infer_types is on, matching what process.env returns. The reverse direction accepts a flat mapping or one of those snippet shapes and rejects nested values instead of inventing nested names. Not a Kubernetes Secret generator and not a shell parser. - **essay_outline_generator** β€” Generate a structured essay outline (introduction, body paragraphs, conclusion) for five essay types and three lengths, and return each section with its guidance points plus Markdown and plain-text renderings. The outline is a deterministic template filled with the topic, thesis, and main points. - **faq_schema_generator** β€” Generate FAQPage JSON-LD structured data from question-answer pairs, and return the JSON-LD string, a ready