RedBison

Search RedBison's 500+ tools by keyword: developer utilities, file and media converters, and finance, health, math and engineering calculators. Returns a ranked list with each tool's canonical URL and, when one exists, its in-depth guide. Use this when a user asks 'is there a tool for X?' or 'what RedBison tool should I use to Y?' Returns: JSON with { query, count, results[] } where each result has { id, name, description, url, category }. Example input: {"query":"json format","limit":5}

Get full metadata for a specific RedBison tool by its id (slug). Returns: Full JSON metadata for the tool: { id, name, description, url, category, subcategory?, keywords[] }. Example input: {"id":"jwt-decoder"}

List all tool categories and subcategories available on RedBison. Returns: JSON array of { category, count } entries covering the full RedBison catalog. Example input: {}

无需输入参数。

Return server status, name, version, and uptime info. Use this to monitor the MCP server. Returns: JSON { status, name, version, startedAt, timestamp }. Example input: {}

无需输入参数。

Base64-encode a UTF-8 string. Returns: Base64-encoded string (UTF-8 safe). Example input: {"text":"hello world"}

Base64-decode a string back to UTF-8. Returns: UTF-8 decoded plain text. Example input: {"text":"aGVsbG8gd29ybGQ="}

Percent-encode a string for use in URLs. Returns: Percent-encoded string suitable for URL query values. Example input: {"text":"hello world & friends"}

Percent-decode a URL-encoded string. Returns: Percent-decoded plain text. Example input: {"text":"hello%20world%20%26%20friends"}

Pretty-print a JSON string with a chosen indent size. Returns: Pretty-printed JSON string with the requested indent, or an error message if the input is invalid. Example input: {"text":"{\"a\":1,\"b\":[2,3]}","indent":2} On error, includes hint: Pass valid JSON as `text` (double-quoted keys/strings, no trailing commas). `indent` is an integer between 0 and 10.

Validate a JSON string. Returns { valid, error? }. Returns: JSON { valid: boolean, error?: string }. Example input: {"text":"{\"a\":1}"} On error, includes hint: Pass the JSON string as `text`.

Parse CSV text into JSON. Uses the first row as headers unless header=false. Returns: JSON array of row objects keyed by the header row. Example input: {"text":"name,age\nAda,36\nGrace,85","delimiter":","} On error, includes hint: `text` must be CSV with a header row unless `header:false`. Escape embedded quotes as "".

Convert an array of JSON objects to CSV. Columns are inferred from the union of keys. Returns: CSV string with a header row derived from the first object's keys. Example input: {"json":"[{\"name\":\"Ada\",\"age\":36},{\"name\":\"Grace\",\"age\":85}]","delimiter":","} On error, includes hint: `json` must be a JSON-encoded array of objects (parameter name is `json`, not `text`).

Convert a JSON string to YAML. Returns: YAML representation of the parsed JSON value. Example input: {"json":"{\"name\":\"Ada\",\"skills\":[\"math\",\"logic\"]}"} On error, includes hint: Pass the JSON as `json` (a stringified JSON value), not as a JS object.

Convert simple YAML (maps, lists, scalars, flow {} / []) to JSON. Anchors, tags, and complex block scalars are not supported. Returns: Pretty-printed JSON built from the YAML input. Example input: {"yaml":"name: Ada\nskills:\n - math\n - logic"} On error, includes hint: Pass the YAML document as `yaml`; only 2-space indented block YAML is supported.

Parse tab-separated values into JSON. First row is treated as header unless header=false. Returns: JSON array of row objects derived from the TSV header row. Example input: {"text":"name\tage\nAda\t36\nGrace\t85"} On error, includes hint: `text` must use real tab (\t) separators, not multiple spaces.

Pretty-print an HTML fragment. Returns: Pretty-printed HTML string with normalized indentation. Example input: {"html":"<div><p>hi</p><p>there</p></div>","indent":2}

Pretty-print an XML document. Returns: Pretty-printed XML string with normalized indentation. Example input: {"xml":"<root><a>1</a><b>2</b></root>","indent":2}

Normalize YAML indentation and whitespace. Returns: Normalized YAML string (2-space indent, collapsed blank lines). Example input: {"yaml":"a: 1\nb: 2\n\n\nc: 3"}

Format a SQL query with keyword uppercasing and line breaks before major clauses. Returns: Formatted SQL query with keywords on new lines. Example input: {"sql":"select id,name from users where active=true order by name"}

Convert text to a URL-safe slug. Returns: URL-safe slug (lowercase, dashes). Example input: {"text":"Hello, World! 2026"}

Convert text between upper, lower, title, camel, pascal, snake, and kebab case. Returns: String converted to the requested case. Example input: {"text":"hello world example","style":"camel"} On error, includes hint: `style` must be one of: upper, lower, title, camel, pascal, snake, kebab.

Escape or unescape a string for JSON, HTML, XML, JavaScript, or SQL. Returns: Escaped or unescaped string for the chosen format. Example input: {"text":"He said \"hi\"","format":"json","mode":"escape"} On error, includes hint: `mode` is 'escape' or 'unescape'; `format` is one of: json, html, xml, js, sql.

Remove duplicate lines from text while preserving first-occurrence order. Returns: Text with duplicate lines removed (order preserved). Example input: {"text":"apple\nbanana\napple\ncherry","caseInsensitive":false,"trim":true}

Return lines that appear more than once in the input, with their occurrence counts. Returns: JSON array of { value, count } for lines appearing more than once. Example input: {"text":"apple\nbanana\napple\ncherry\nbanana"}

Collapse runs of whitespace inside each line to a single space and trim ends. Returns: Text with runs of whitespace collapsed to a single space. Example input: {"text":"hello world foo"}

Strip blank / whitespace-only lines from the input. Returns: Text with blank lines removed. Example input: {"text":"one\n\ntwo\n\n\nthree"}

Join lines with a separator (comma by default). Empty lines are skipped. Returns: Single-line string with items joined by the chosen separator. Example input: {"text":"apple\nbanana\ncherry","separator":", ","quote":false}

Split a delimited string into one item per line. Returns: Multi-line string, one item per line. Example input: {"text":"apple, banana, cherry","separator":",","trim":true}

Decode a JWT and return its header + payload. Does NOT verify the signature. Returns: JSON with decoded { header, payload, signature } sections of the JWT. Example input: {"token":"eyJhbGciOiJIUzI1NiJ9.eyJzdWIiOiIxMjMifQ.abc"} On error, includes hint: `token` must be a compact JWT with three base64url-encoded parts separated by dots.

Compute a cryptographic hash of a string. Returns: Hex digest of the input using the requested algorithm. Example input: {"text":"hello world","algorithm":"sha256"} On error, includes hint: `algorithm` must be one of: md5, sha1, sha256, sha384, sha512.

Hash two inputs with the same algorithm and compare them in constant time. Useful for verifying file integrity. Returns: JSON { match: boolean } computed with a constant-time comparison. Example input: {"a":"hello","b":"hello","algorithm":"sha256"} On error, includes hint: `algorithm` (optional, default sha256) must be one of: md5, sha1, sha256, sha384, sha512.

Generate one or more v4 UUIDs. Returns: JSON array of v4 UUID strings. Example input: {"count":3} On error, includes hint: `count` must be an integer between 1 and 100.

Generate a cryptographically random password. Choose length and character classes. Returns: Randomly generated password string. Example input: {"length":20,"uppercase":true,"lowercase":true,"digits":true,"symbols":true,"count":1} On error, includes hint: `length` 4-256. Enable at least one character class (lowercase/uppercase/digits/symbols).

Hash a password with bcrypt. Returns: Bcrypt hash string (includes cost + salt). Example input: {"password":"s3cret!","cost":10} On error, includes hint: `cost` must be an integer between 4 and 12. Higher cost is exponentially slower.

Check whether a password matches a bcrypt hash. Returns: JSON { match: boolean }. Example input: {"password":"s3cret!","hash":"$2a$10$abcdefghijklmnopqrstuv"} On error, includes hint: `hash` must be a full bcrypt string starting with $2a$, $2b$, or $2y$.

Given a CIDR block (e.g. 10.0.0.0/24), return network, broadcast, netmask, first/last host, and host count. Returns: JSON with network, broadcast, first/last host, host count, and mask for the CIDR. Example input: {"cidr":"192.168.1.0/24"} On error, includes hint: `cidr` must be IPv4 in `A.B.C.D/prefix` form, prefix 0-32.

Convert an IPv4 address between dotted-decimal, integer, binary, and hex forms. Accepts any of those as input. Returns: JSON { decimal, integer, hex, binary } representations of the IPv4 address. Example input: {"value":"192.168.1.1"} On error, includes hint: `value` accepts dotted (192.168.0.1), integer (3232235521), 32-bit binary, or 0x-prefixed hex.

Convert between Unix epoch seconds/milliseconds and ISO 8601. Pass a number to convert to ISO, or an ISO string to convert to epoch. Returns: JSON with ISO string, epoch seconds, epoch milliseconds, and a UTC-formatted representation. Example input: {"value":"1700000000"} On error, includes hint: `value` is a string: either an integer (seconds or ms since epoch) or an ISO 8601 timestamp.

Convert a timestamp between two IANA timezones. Returns the wall-clock time in the target zone. Returns: JSON { utc, zone, local } with the timestamp rendered in the target IANA timezone. Example input: {"timestamp":"2026-07-11T12:00:00Z","to":"America/Los_Angeles"} On error, includes hint: `timestamp` must be ISO 8601 (e.g. 2026-07-11T12:00:00Z); `to` must be an IANA zone like America/New_York.

Compute the difference between two ISO 8601 timestamps in days, hours, minutes, and seconds. Returns: JSON with milliseconds, totalSeconds/Minutes/Hours/Days, and a { days, hours, minutes, seconds } breakdown. Example input: {"from":"2026-01-01T00:00:00Z","to":"2026-07-11T12:00:00Z"} On error, includes hint: Both `from` and `to` must be ISO 8601 timestamps.

Validate a standard 5-field cron expression (minute hour day month weekday). Returns a human-readable summary and any parse error. Returns: JSON { valid: boolean, nextRuns?: string[], error?: string }. Example input: {"expression":"*/5 * * * *"} On error, includes hint: `expression` must be a 5-field cron (minute hour dom month dow). 6-field / Quartz syntax is not supported.

Test a regular expression against text. Returns all matches and capture groups. Returns: JSON { matches: string[], count } of matches for the pattern. Example input: {"pattern":"\\d+","flags":"g","text":"order 42 shipped on day 7"} On error, includes hint: `pattern` is a JS RegExp source (double-escape backslashes in JSON). `flags` uses standard flags like g, i, m, s.

Compute a line-level diff between two strings. Returns a unified-style patch and per-line change list. Returns: JSON line-diff with { added, removed, unchanged } arrays. Example input: {"a":"line 1\nline 2\nline 3","b":"line 1\nline 2 changed\nline 3"}

Compare two newline-separated lists. Returns union, intersection, only-in-A, and only-in-B. Returns: JSON { union, intersection, onlyInA, onlyInB } arrays. Example input: {"a":"apple\nbanana\ncherry","b":"banana\ncherry\ndate"}

Convert a color between hex, rgb, and hsl. Accepts inputs like '#3B82F6', 'rgb(59,130,246)', or 'hsl(217,91%,60%)'. Returns: JSON with hex, rgb, and hsl representations of the color. Example input: {"color":"#3b82f6"} On error, includes hint: `color` accepts #rgb, #rrggbb, rgb(r,g,b), or hsl(h,s%,l%).

Compute the WCAG contrast ratio between two hex colors and report AA/AAA pass/fail for normal and large text. Returns: JSON { ratio, AA, AAA, AALarge, AAALarge } for the two colors. Example input: {"foreground":"#ffffff","background":"#3b82f6"} On error, includes hint: Both colors must be hex (#rgb or #rrggbb).

Build a CSS linear or radial gradient string from color stops. Returns: CSS gradient string (linear or radial). Example input: {"type":"linear","angle":45,"stops":["#3b82f6","#8b5cf6"]} On error, includes hint: `type` is 'linear' or 'radial'; provide at least 2 color `stops`.

Generate random human names. Returns: JSON array of random human name strings. Example input: {"count":3}

Generate random-looking (non-deliverable) test email addresses. Returns: JSON array of random email address strings. Example input: {"count":3,"domain":"example.com"}

Generate cryptographically random numbers in a range. Returns: JSON array of generated numbers. Example input: {"min":1,"max":100,"count":5,"mode":"integer"}

Generate placeholder lorem ipsum text. Returns: Lorem ipsum text of the requested size. Example input: {"unit":"paragraphs","count":2}

Compute the equated monthly installment (EMI) for a loan. Returns: JSON { emi, totalInterest, totalPayment } for the loan. Example input: {"principal":500000,"annualRatePercent":8.5,"months":240}

Compute simple interest: I = P × r × t. Returns: JSON { interest, total } computed as P·r·t. Example input: {"principal":10000,"annualRatePercent":5,"years":3}

Compute compound interest: A = P(1 + r/n)^(nt). Returns: JSON { interest, total } computed as P(1+r/n)^(nt). Example input: {"principal":10000,"annualRatePercent":5,"years":10,"compoundsPerYear":12}

Calculate Body Mass Index in metric (kg/cm) or imperial (lb/in) units. Returns: JSON { bmi, category } (Underweight / Normal / Overweight / Obese). Example input: {"weight":70,"height":175,"units":"metric"}

List the calculators that run_calculator can execute (finance, health, maths, statistics, construction, science, everyday and engineering), with each one's inputs, units, defaults and an example. Filter with an optional search query. Returns: JSON { total, calculators[] } where each has { id, title, summary, url, inputs (JSON Schema per field), example }. Example input: {"query":"mortgage","limit":5}

Run any RedBison calculator by id with named inputs and get the result as JSON: mortgages and loans, savings and investing, pay and tax, body and fitness, maths, statistics, geometry, construction, science and engineering (beams, bolts, wire size, PCB traces, pipe flow, rafters, sheet metal and more). Call list_calculators first to see ids, inputs and units. Returns: JSON { id, title, url, inputs, result } with the calculator's outputs. Example input: {"id":"percentage-calculator","inputs":{}}

Render text as a FIGlet ASCII art banner in one of 7 bundled fonts (Standard, Slant, Small, Big, Doom, ANSI Shadow, Mini), optionally wrapped at a column width and inside code comment markers. Printable ASCII only. The page at redbison.dev/ascii-art-text-generator has more fonts. Returns: Text: the banner (show it in a monospace code block). structuredContent { art, font, lines, columns, unsupported[] }. Example input: {"text":"Hello","font":"Slant","layout":"default","width":80,"comment":"none"}

Compute or validate the GS1 mod-10 check digit of an EAN-8, UPC-A, EAN-13 or ITF-14 (GTIN-8/12/13/14) number. Give the digits without the check digit to compute it, or the full code to validate it. Returns: JSON { digits, checkDigit, fullCode, valid, gtin14, formats[] }. valid is true/false when a full code was checked and null when the check digit was computed. Example input: {"code":"400638133393","mode":"compute"}

Convert Unix permissions between octal (755, 4755), symbolic (rwxr-xr-x) and ls -l strings, or apply a chmod expression such as 'go-w,u+x' to a starting mode. Explains each class, setuid/setgid/sticky bits and risky modes, and gives umask defaults. Returns: JSON { octal, symbolic, ls, description[], warnings[], command, symbolicArg, umaskDefaults{ umask, newFile, newDirectory } }. Example input: {"mode":"rwxr-x---"} On error, includes hint: Use 1-4 octal digits, nine rwx characters (optionally with a leading file type), or a symbolic expression like u+x,go-w.

Build a harmonious color palette from one base color (complementary, analogous, triadic, split-complementary, tetradic, square, monochromatic or tints and shades). Returns each color as HEX, RGB, HSL and OKLCH with WCAG contrast against white and black, an 11-step 50-950 scale of the base, and ready-to-paste CSS variables and Tailwind config. Same maths as https://redbison.dev/color-palette-generator. Returns: JSON { base, rule, wheel, colors[] { role, hex, rgb, hsl, oklch, contrastOnWhite, contrastOnBlack, bestTextColor, wcagNormalText }, scale { '50'..'950': hex }, baseScaleStep, css, tailwindV3, tailwindV4 }. Example input: {"base":"#3b82f6","rule":"triadic","wheel":"rgb","name":"brand"} On error, includes hint: Pass the base color as hex like #3b82f6 (rgb(), hsl(), oklch() and CSS names also work). rule must be one of complementary, analogous, triadic, split-complementary, tetradic, square, monochromatic, shades.

Build a standard five-field cron expression from plain choices (every N minutes, hourly, daily, weekly, monthly, yearly) or explain an existing one. Returns the expression, an English description, the next run times in a time zone, and the same schedule for crontab, Kubernetes, GitHub Actions, Quartz and AWS EventBridge. Returns: JSON { expression, description, nextRuns[], timezone, platforms[{ id, name, value, notes[] }], warnings[] }. value is null on Quartz/EventBridge when both day fields are restricted. Example input: {"frequency":"weekly","daysOfWeek":["mon","wed","fri"],"hours":[9],"minute":30,"timezone":"America/New_York"} On error, includes hint: Give either plain choices (frequency plus minute/hours/days) or a five-field expression. Days of week accept names or 0-6 with 0 = Sunday.

Pretty-print or minify CSS, SCSS or Less, optionally sorting declarations alphabetically. Sorting is cascade-safe: a shorthand and its longhands (border / border-top, margin / margin-top, font / line-height) keep their original relative order. Minify removes comments and whitespace only; values such as 0px or colours are never rewritten. Returns: Text: the stylesheet. structuredContent { code, movedDeclarations, warnings[] }. Example input: {"css":".a{z-index:1;color:red;border-top:1px solid;border:0}","mode":"beautify","syntax":"css","sort":true,"indent":2} On error, includes hint: Unbalanced braces or an unterminated string or comment are reported as errors.

Convert a curl command (bash/zsh quoting, $'...' strings, \ line continuations, or Windows cmd ^ escaping from 'Copy as cURL') into JavaScript fetch, Node Axios, Python requests, Go net/http or PHP ext-curl. Follows curl semantics: -d implies POST with application/x-www-form-urlencoded, repeated -d joins with &, --json sets JSON headers, -G moves data to the query string, -F is multipart, -u is Basic auth, no redirects without -L. Nothing is sent over the network. Returns: Text: the generated code, followed by notes on anything not translated. structuredContent { code, method, url, headers[{ name, value }], hasBody, warnings[] }. Example input: {"command":"curl -X POST https://api.example.com/v1/items -H 'Content-Type: application/json' -d '{\"name\":\"Ada\"}'","target":"python"} On error, includes hint: Send the command exactly as copied, including quotes. The word curl must be present.

Convert one or more 'docker run' commands into a Compose file (services, ports, volumes, environment, networks, healthcheck, GPUs, limits), or convert a compose.yaml back into docker run commands. Unsupported flags are reported as warnings. Returns: run_to_compose: JSON { compose (YAML text), services[], warnings[] }. compose_to_run: JSON { commands[], warnings[] }. Example input: {"input":"docker run -d --name web -p 8080:80 -v html:/usr/share/nginx/html --restart unless-stopped nginx:alpine"} On error, includes hint: Each command must start with 'docker run' (or 'docker container run'). Separate several commands with new lines.

Combine .gitignore templates (based on github/gitignore) for languages, frameworks, tools, editors and operating systems into one deduplicated file, optionally test paths against it. Template ids: node, python, django, jupyter, java, maven, gradle, kotlin, android, go, rust, ruby, rails, php, laravel, dotnet, c, cpp, swift, flutter, elixir, scala, r, nextjs, terraform, unity, docker, secrets, vscode, jetbrains, visualstudio, eclipse, vim, emacs, sublime, macos, windows, linux. Returns: JSON { content (.gitignore text), unknownTemplates[], duplicatesRemoved, tests[{ path, ignored, rule{pattern,line,negate}, via }] }. Example input: {"templates":["node","macos","vscode"],"testPaths":[".env","node_modules/x.js",".vscode/settings.json"]}

Compute an HMAC (RFC 2104) with SHA-256, SHA-1, SHA-384 or SHA-512 and return it in hex, Base64 and Base64url. Optionally build a GitHub, Stripe or Slack webhook signature, and check an expected signature with a constant-time comparison. Returns: JSON { algorithm, hex, base64, base64url, bits, key_bytes, message_bytes, signed_string, header, match } where match is null unless `expected` was given. Example input: {"message":"The quick brown fox jumps over the lazy dog","key":"key","algorithm":"SHA-256"} On error, includes hint: `algorithm` is SHA-256, SHA-1, SHA-384 or SHA-512; `key_encoding` and `message_encoding` are utf8, hex or base64; `preset` is plain, github, stripe or slack.

Look up HTTP status codes: meaning, defining RFC section (RFC 9110 and others from the IANA registry), when to send it and common pitfalls. Pass a code (404), a class (5xx) or words (rate limit). Unofficial nginx and Cloudflare codes are included and flagged. Returns: JSON { query, count, results[] } where each result has { code, name, class, spec, official, deprecated, cacheableByDefault, summary, usage, note }, plus unknownCode text when a 3-digit code is not registered. Example input: {"query":"429"}

Format JavaScript, TypeScript or JSX with js-beautify, or minify JavaScript/TypeScript by removing comments and whitespace (no renaming, no rewriting; line breaks are kept where automatic semicolon insertion needs them). The output is re-lexed and compared token by token with the input. Returns: Text: the formatted or minified code. structuredContent { code, mode, inputBytes, outputBytes, savedPercent, warnings[] }. Example input: {"code":"function add(a,b){return a+b}","mode":"beautify","language":"javascript","indent":2,"brace_style":"collapse","keep_license":true} On error, includes hint: Pass the whole file as `code`. An unterminated string, comment or regex is reported with its line and column.

Validate a JSON document against a JSON Schema (drafts 06, 07, 2019-09 and 2020-12, via Ajv, all errors reported with JSON Pointer paths and formats such as date-time, email, uuid, ipv4/ipv6 checked), or infer a starter schema from sample JSON. Leave `schema` empty to infer. Returns: Validate: text summary plus structuredContent { mode, valid, draft, errors[{ path, message, keyword, schemaPath, params }], warnings[] }. Infer: the schema as JSON text and structuredContent { mode, schema }. Example input: {"data":"{\"name\":\"Ada\",\"age\":-1}","schema":"{\"type\":\"object\",\"properties\":{\"age\":{\"type\":\"integer\",\"minimum\":0}},\"required\":[\"name\",\"email\"]}","draft":"auto"} On error, includes hint: Both `data` and `schema` are JSON text, not objects. Send an empty `schema` to generate one from `data`.

Turn CSV, TSV (spreadsheet paste) or an existing Markdown table into a GitHub Flavored Markdown table. The first row is the header; pipes are escaped as \|, line breaks inside cells become <br>, and columns can be padded so the pipes line up (CJK-width aware). Returns: Text: the Markdown table. structuredContent { markdown, rows (including header), columns, detected ('csv' | 'tsv' | 'markdown' | 'other'), errors[] }. Example input: {"text":"Item,Qty,Price\nApple,3,$1.20\nKiwi,12,$0.50","align":"none","pretty":true,"outer_pipes":true,"bold_header":false} On error, includes hint: Put the header in the first row. Quote CSV fields that contain commas or line breaks.

Find the MIME (media) type for a file extension or file name, or the extensions for a MIME type. Returns the type, whether it is IANA-registered, aliases, notes and the Content-Type header to send. Returns: JSON { query, matchedBy: 'extension'|'mime'|'text', count, results[] } with { mime, extensions[], name, category, registered, aliases[], spec, note, contentType }. Example input: {"query":"woff2"}

Translate text to International Morse code (ITU-R M.1677-1) or Morse back to text. Letters are separated by spaces and words by ' / '. Wrap prosigns in angle brackets (<SOS>, <AR>, <SK>). Also returns the transmission time at a given speed in words per minute, with optional Farnsworth spacing. Same rules as redbison.dev/morse-code-translator. Returns: Text: the Morse or decoded text. structuredContent { direction, morse | text, unsupported[] | unknownCodes[], durationSeconds, unitMs }. Example input: {"input":"SOS at 18 wpm","direction":"encode","wpm":18}

Spell text with the ICAO/NATO radiotelephony alphabet (Alfa, Bravo, Charlie...), with ICAO number words (Tree, Fower, Fife, Niner) or plain ones, optional Capital/Lowercase markers for passwords, or decode code words back to text. Same rules as redbison.dev/nato-phonetic-alphabet. Returns: Text: code words, ' / ' between input words. structuredContent { spelled, tokens[{char, word, pronunciation}] } or, for decode, { text, unknown[] }. Example input: {"text":"Gate B12","direction":"spell","spelling":"icao","digits":"icao","markCase":false}

Estimate how guessable a password is with zxcvbn pattern matching (dictionary words, common passwords, keyboard walks, sequences, dates, l33t). Returns a 0-4 score, estimated guesses, entropy in bits, crack times for online and offline attacks, warnings, suggestions and NIST SP 800-63B-4 length checks. The password is not echoed back. Returns: JSON { score 0-4, label, length, guesses, guesses_log10, entropy_bits, character_pool_entropy_bits, character_classes[], crack_times[{scenario, seconds, display}], warning, suggestions[], patterns[{pattern, start, end, description}], nist_sp800_63b{...} }. The password itself is never included. Example input: {"password":"Tr0ub4dor&3","user_inputs":["alice"]} On error, includes hint: `password` must be a string of at most 256 characters; `user_inputs` is an optional array of strings.

Score English text with Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog, SMOG, Coleman-Liau and the Automated Readability Index, and list hard sentences, passive voice, -ly adverbs and wordy phrases with plainer alternatives. Syllables are estimated with rules plus an exception list. Same rules as redbison.dev/readability-checker. Returns: JSON { counts, wordsPerSentence, syllablesPerWord, fleschReadingEase, fleschBand, fleschKincaidGrade, gunningFog, smog, smogShortSample, colemanLiau, automatedReadabilityIndex, averageGrade, hardSentences, veryHardSentences, passiveVoice[], adverbs[], plainerWords[{found,try}], hardest[{sentence,words,grade,level}] }. Example input: {"text":"The committee utilized a methodology that was developed in order to facilitate a comprehensive assessment of the program. We read it twice.","lineBreaksEndSentences":true,"maxSentences":5}

Generate 301/302/307/308 redirect rules for Apache .htaccess (mod_rewrite), nginx, Netlify _redirects or vercel.json from a list of 'source destination [code]' lines. A source ending in /* moves a whole folder. Also checks for loops, chains and duplicates and can add HTTPS and www/non-www rules. Returns: JSON { platform, config (file text), ruleCount, issues[{ line, level, message }], notes[] }. Example input: {"rules":"/old-page /new-page\n/blog/* /news/:splat 301","platform":"nginx"}

Convert between TOML 1.0, JSON and YAML, validating the input. TOML dates become RFC 3339 strings in JSON; nulls are dropped when writing TOML (it has no null) and reported; integers beyond 2^53 are kept exact; inf and nan become strings in JSON. Use from=toml,to=toml to validate and normalise a TOML file. Returns: Text: the converted document. structuredContent { output, warnings[] } where warnings list dropped nulls, large integers and non-finite numbers. Example input: {"text":"[package]\nname = \"demo\"\nversion = \"0.1.0\"\n\n[dependencies]\nserde = \"1.0\"","from":"toml","to":"json","indent":2} On error, includes hint: TOML output needs an object at the top level; wrap arrays in a key.

Parse a User-Agent header into browser, engine, OS, device type, vendor, model and CPU (ua-parser-js), and detect search, AI, social, SEO and monitoring bots and HTTP libraries. Notes explain User-Agent reduction limits such as Windows 11 reporting as Windows NT 10.0. Returns: JSON { summary, browser{name,version,major}, engine{name,version}, os{name,version}, device{type,vendor,model}, cpu{architecture}, isBot, bot{name,category,owner}|null, notes[] }. Example input: {"userAgent":"Mozilla/5.0 (iPhone; CPU iPhone OS 17_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Mobile/15E148 Safari/604.1"}

Add Google Analytics UTM parameters (utm_source, utm_medium, utm_campaign, utm_id, utm_term, utm_content) to a landing page URL. Keeps existing query parameters and the #fragment, lowercases values by default, and predicts the GA4 default channel group. Returns: JSON { url, params, predictedChannel, warnings[] }: the tagged URL, the UTM values used, the likely GA4 default channel group (approximation) and any naming warnings. Example input: {"url":"https://example.com/pricing?plan=pro","source":"newsletter","medium":"email","campaign":"October Launch","content":"hero_button"}

Count words, characters (with and without spaces), sentences, paragraphs and lines in text, estimate reading and speaking time, list the most frequent words and check the text against common platform limits (X post weighted length, Bluesky, LinkedIn, SMS segments, meta description). Same rules as redbison.dev/word-counter. Returns: JSON { words, characters, charactersNoSpaces, sentences, paragraphs, lines, uniqueWords, utf8Bytes, averageWordLength, averageSentenceLength, readingTime, readingSeconds, speakingTime, speakingSeconds, topWords[{word,count,percent}], platformLimits[{platform,limit,used,remaining,over,detail}] }. Example input: {"text":"Short sentences help. Long ones, full of clauses that wander, do not.","readingWpm":238,"speakingWpm":150,"topWords":5}

Explain how to use any RedBison MCP tool, with its input schema, a sample input and the sample output it produces. Call with no arguments for the list of tools; pass `tool` for one tool, and `calculator` with tool 'run_calculator' for one calculator. Returns: JSON with a tool's { usage, returns, inputSchema, sampleInput, sampleOutput }, or the list of tools when called with no arguments. Example input: {"tool":"list_calculators"}