pillowthought api — every text tool over plain HTTP. text in, text out. usage curl -s https://api.pillowthought.xyz/lorem curl -s "https://api.pillowthought.xyz/base64/encode?in=hello" echo $TOKEN | curl -s --data-binary @- https://api.pillowthought.xyz/jwt/decode curl -s --data-binary @notes.txt https://api.pillowthought.xyz/detect input the POST body (use --data-binary; plain -d strips newlines), or ?in=…, or the rest of the path: https://api.pillowthought.xyz/base64/encode/hello one trailing newline is dropped so `echo` works; add ?raw=1 to keep it. up to 256 KB of text. options as query parameters, e.g. ?count=5. unknown options are an error. output plain text. add ?json=1 for {"tool","output"}; errors come back as plain text with a 4xx status. help https://api.pillowthought.xyz/help/ for one tool's options and an example. tools (104) /base64/decode Decode Base64 (standard or URL-safe) to text. /base64/encode Encode text as Base64 (standard or URL-safe). ?urlSafe=true|false /binary/decode Decode 8-bit binary groups (01001000 01101001) into text. /binary/encode UTF-8 bytes of text as 8-bit binary groups. /case/camel Convert text or identifiers to camelCase (helloWorldAgain), line by line. /case/constant Convert text or identifiers to CONSTANT_CASE (HELLO_WORLD_AGAIN), line by line. /case/dot Convert text or identifiers to dot.case (hello.world.again), line by line. /case/kebab Convert text or identifiers to kebab-case (hello-world-again), line by line. /case/lower Convert text or identifiers to lower case (hello world again), line by line. /case/pascal Convert text or identifiers to PascalCase (HelloWorldAgain), line by line. /case/path Convert text or identifiers to path/case (hello/world/again), line by line. /case/sentence Convert text or identifiers to Sentence case (Hello world again), line by line. /case/snake Convert text or identifiers to snake_case (hello_world_again), line by line. /case/title Convert text or identifiers to Title Case (Hello World Again), line by line. /case/upper Convert text or identifiers to UPPER CASE (HELLO WORLD AGAIN), line by line. /chmod Convert Unix permissions between 755, rwxr-xr-x and u=rwx,go=rx, with a plain-English reading. /color/convert Convert a colour between hex, rgb(), hsl(), hwb(), oklch(), oklab() and lab(). /cron/explain Explain a cron expression in plain English and list the next runs. ?tz=UTC&count=0..50 /css/format Beautify CSS with consistent indentation. ?indent=2|4|tab /css/minify Strip comments and whitespace from CSS. /csv/to-json Convert CSV/TSV (first row = headers) to a JSON array of objects. ?header=true|false&types=true|false&delimiter=auto|,|tab|;|| /domain/parse Split a hostname into subdomain, registrable domain and public suffix; convert IDN ↔ punycode. /duration/convert Convert durations between 1h30m, ISO 8601 (PT1H30M), hh:mm:ss, seconds and plain English. /email/parse Validate an email address and split it into name, local part, +tag and domain. /escape/csv Escape text for use inside a CSV field. /escape/js Escape text for use inside a JavaScript string. /escape/json Escape text for use inside a JSON string. /escape/regex Escape text for use inside a regular expression. /escape/shell Escape text for use inside a shell argument. /escape/sql Escape text for use inside a SQL string literal. /hash/crc32 CRC-32 checksum digest of UTF-8 text. ?encoding=hex|HEX|base64|base64url /hash/identify Guess which algorithm produced a hash from its length and format. /hash/md5 MD5 digest of UTF-8 text. Not for security — collisions are easy. ?encoding=hex|HEX|base64|base64url /hash/sha1 SHA-1 digest of UTF-8 text. Not collision-resistant any more. ?encoding=hex|HEX|base64|base64url /hash/sha256 SHA-256 digest of UTF-8 text. ?encoding=hex|HEX|base64|base64url /hash/sha384 SHA-384 digest of UTF-8 text. ?encoding=hex|HEX|base64|base64url /hash/sha512 SHA-512 digest of UTF-8 text. ?encoding=hex|HEX|base64|base64url /hex/decode Decode hexadecimal bytes back into UTF-8 text. /hex/encode UTF-8 bytes of text as hexadecimal. ?sep=none|space|colon|0x|\x&upper=true|false /hmac/sha256 Keyed HMAC signature of text (SHA-256 by default; SHA-1/384/512 too). ?key=…&algo=sha256|sha512|sha384|sha1&encoding=hex|HEX|base64|base64url /html/decode Turn < © é and friends back into characters. /html/encode Escape <, >, &, quotes (and optionally all non-ASCII) as HTML entities. ?all=true|false /html/format Beautify HTML with consistent indentation. ?indent=2|4|tab /html/minify Collapse whitespace and drop comments in HTML (keeps

,  and <script> intact).
  /http/status        What an HTTP status code means: 404, 418, 502…
  /ip/info            IPv4 and IPv6: private or public, reserved ranges, and subnet maths for CIDR blocks.
  /js/format          Beautify JS with consistent indentation.
                      ?indent=2|4|tab
  /json/format        Pretty-print JSON, with a clear line/column message when it's invalid.
                      ?indent=2|4|tab&sort=true|false
  /json/minify        Remove all whitespace from JSON.
  /json/query         Run a jq filter (.items[] | select(.price < 10)) or a JSONPath ($.items[*].name) against JSON.
                      ?query=.&mode=auto|jq|jsonpath&raw=true|false&compact=true|false
  /json/sort-keys     Sort object keys alphabetically, recursively.
                      ?indent=2|4|tab
  /json/to-csv        Convert an array of JSON objects to CSV (nested keys become dot.columns).
                      ?delimiter=,|tab|;
  /json/to-sql        Turn an array of JSON objects into CREATE TABLE + INSERT statements.
                      ?table=items&dialect=postgres|mysql|sqlite&create=true|false
  /json/to-toml       Convert a JSON object to TOML.
  /json/to-ts         Generate TypeScript interfaces from a JSON sample (merges array items, marks optional fields).
                      ?root=Root&type=true|false
  /json/to-xml        Convert JSON to XML. Keys starting with @_ become attributes; #text is element text.
                      ?root=root
  /json/to-yaml       Convert JSON to YAML.
                      ?indent=2|4
  /json/validate      Check that JSON is valid; explains the first problem with its line and column.
  /jwt/decode         Decode a JSON Web Token's header and payload and show when it expires. Doesn't verify the signature.
  /lines/dedupe       Remove repeated lines, keeping the first occurrence and the original order.
                      ?ci=true|false&trim=true|false
  /lines/reverse      Reverse the order of lines.
  /lines/shuffle      Shuffle lines into a random order (cryptographically random).
  /lines/sort         Sort lines alphabetically, naturally, numerically or by length.
                      ?order=asc|desc&mode=natural|text|numeric|length&ci=true|false
  /lines/trim         Trim whitespace from each line and drop empty lines.
                      ?empty=true|false&collapse=true|false
  /lorem              Placeholder text: words, sentences or paragraphs. Input is ignored.
                      ?unit=paragraphs|sentences|words&count=1..500&classic=true|false
  /md/to-html         Render Markdown (GitHub-flavoured) to HTML. Output isn't sanitised — don't inject untrusted HTML.
  /mime/lookup        File extension → MIME type, or MIME type → extensions.
  /nanoid             Generate URL-friendly Nano IDs (21 characters by default). Input is ignored.
                      ?size=2..256&alphabet=…&count=1..1000
  /number/base        Convert integers between binary, octal, decimal, hex, base 36 and any base 2–36 (arbitrary size).
                      ?from=auto|2|8|10|16|36&to=0..36
  /password           Strong random passwords with letters, digits and symbols. Input is ignored.
                      ?length=4..256&lower=true|false&upper=true|false&digits=true|false&symbols=true|false&ambiguous=true|false&count=1..100
  /password/strength  Estimate a password's entropy and how long an offline attack would take. Nothing is sent anywhere.
  /pem/decode         Read a PEM certificate (or chain, CSR, public key): subject, issuer, validity, SANs, key, fingerprints.
  /random/string      Random strings from a chosen character set (crypto-strength). Input is ignored.
                      ?length=1..4096&charset=alnum|lower|letters|hex|numeric|base58|crockford&count=1..1000
  /regex/explain      Break a regular expression into its parts and say what each one matches (offline, no AI).
  /semver/compare     Compare two versions (1.2.3 vs 1.10.0), or check a version against a range like ^1.2 || >=2.1.
  /semver/parse       Split a semantic version into major, minor, patch, pre-release and build, and show the next versions.
  /slug               Turn a title into a URL slug: “Héllo, Wörld!” → hello-world.
                      ?sep=-|_|.&lower=true|false&max=0..500
  /sql/format         Pretty-print SQL for PostgreSQL, MySQL, SQLite, SQL Server, BigQuery and more.
                      ?dialect=sql|postgresql|mysql|mariadb|sqlite|transactsql|plsql|bigquery|snowflake|redshift|spark|duckdb|trino|clickhouse&keywordCase=upper|lower|preserve&indent=2|4|tab
  /svg/to-data-uri    Turn SVG markup into a compact data: URI for CSS backgrounds and <img src>.
                      ?base64=true|false
  /text/count         Count characters, words, sentences, paragraphs, lines and bytes; estimate reading time.
  /time/from-unix     Convert a Unix timestamp (seconds, ms, µs or ns — detected) to readable dates in any time zone.
                      ?tz=UTC&unit=auto|s|ms|us|ns
  /time/now           The current time as Unix seconds, milliseconds, ISO 8601 and in a time zone. Input is ignored.
                      ?tz=UTC
  /time/parse         Read any date and show it as Unix time, ISO 8601, RFC 2822, in a time zone and relative to now.
                      ?tz=UTC
  /time/to-unix       Convert a date (ISO, RFC 2822, “2 hours ago”…) to a Unix timestamp. Times without an offset use the chosen zone.
                      ?tz=UTC&unit=s|ms
  /toml/to-json       Convert TOML (Cargo.toml, pyproject.toml…) to JSON.
                      ?indent=2|4
  /ua/parse           Read the browser, engine, OS, device and CPU out of a User-Agent string.
  /ulid               Generate ULIDs: sortable, time-prefixed 26-character IDs. Input is ignored.
                      ?count=1..1000&lower=true|false
  /ulid/parse         Read the creation time out of a ULID.
  /unescape/csv       Undo CSV field escaping.
  /unescape/js        Undo JavaScript string escaping.
  /unescape/json      Undo JSON string escaping.
  /unescape/regex     Undo regular expression escaping.
  /unescape/shell     Undo shell argument escaping.
  /unescape/sql       Undo SQL string literal escaping.
  /unicode/escape     Escape non-ASCII characters as \uXXXX, \u{…}, &#x…;, U+ code points and more.
                      ?style=js|es6|python|html|css|codepoint&ascii=true|false
  /unicode/unescape   Turn \uXXXX, \u{…}, \xXX, U+XXXX and &#…; back into characters.
  /url/decode         Decode percent-encoded (%20) text.
                      ?plus=true|false
  /url/encode         Percent-encode text for use in a URL.
                      ?mode=component|uri|form
  /url/parse          Split a URL into scheme, host, port, path, query parameters and fragment.
  /uuid               Generate UUIDs (v4 random or v7 time-ordered). Input is ignored.
                      ?version=v4|v7&count=1..1000&upper=true|false
  /uuid/parse         Show a UUID's version, variant and (for v1/v6/v7) when it was created.
  /xml/format         Pretty-print (or validate) XML.
                      ?indent=2|4|tab
  /xml/to-json        Convert XML to JSON. Attributes become @_name keys.
                      ?indent=2|4
  /yaml/to-json       Convert YAML (including multi-document files) to JSON.
                      ?indent=2|4

more
  /detect             what is this text? ranks the tools that recognise it.
  /chain              run several tools in a row:
                      curl -s https://api.pillowthought.xyz/chain -H 'content-type: application/json' \
                        -d '{"steps":[{"id":"base64.decode"},{"id":"jwt.decode"}],"input":"…"}'
                      or: curl -s --data-binary @in.txt "https://api.pillowthought.xyz/chain?steps=base64.decode,jwt.decode"
  /mcp                MCP server (streamable HTTP) with the same tools:
                      claude mcp add --transport http pillowthought https://api.pillowthought.xyz/mcp

not here
  bcrypt.hash, bcrypt.check
  these need more CPU than the free edge allows (10 ms) or a real browser,
  so they only run in your browser on the website.

docs, MCP setup and the browser extension: open https://api.pillowthought.xyz in a browser.