Changelog
New tools, SDK updates, and platform improvements.
September 26, 2026New Tools
v0.71.0 / v0.71.0 / v0.71.0MIDI File Info Reader + Dockerfile Explainer (AI)
- MIDI File Info Reader: parse a Standard MIDI File's header and every track chunk directly from its bytes to report format (single-track, multi-track synchronous, or multi-track asynchronous), tick division, initial tempo, time signature, and a per-track event/note-on count, plus total duration in seconds that accounts for tempo changes across the file. Nothing is played back or rendered to audio. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/midi-file-info`), Node SDK (`readMidiInfo`), Python SDK (`read_midi_info`), and MCP server (`read_midi_info`)
- Dockerfile Explainer (AI): paste an existing Dockerfile and get back a plain-English, instruction-by-instruction walkthrough of what the image build does, plus a list of concrete issues worth fixing, like an unpinned base image tag, running as root, or a package-manager install that never cleans its cache. The reverse of the existing Dockerfile Generator, which writes a new Dockerfile from a description instead of explaining one you already have. Calls GPT-4o mini on the server. Shipped on browser UI and REST API (`POST /v1/ai/dockerfile-explain`); not available via SDK or MCP, matching the site's other AI-generation tools
- Node SDK bumped to v0.71.0, Python SDK to v0.71.0, MCP server to v0.71.0 (276 tools total, up from 274)
September 25, 2026New Tools
v0.70.0 / v0.70.0 / v0.70.0TOTP / HOTP Code Generator + Tax-Equivalent Yield Calculator
- TOTP / HOTP Code Generator: generate the current TOTP code (RFC 6238) from a base32-encoded shared secret, with a live countdown until it rotates, or an HOTP code (RFC 4226) for a specific counter value, supporting SHA-1/SHA-256/SHA-512 and 6-8 digit codes. Useful for testing a 2FA integration without reaching for a phone; the secret is never stored. The browser tool runs the HMAC computation locally via the Web Crypto API. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/totp-hotp-generator`), Node SDK (`generateHotp`, `generateTotp`), Python SDK (`generate_hotp`, `generate_totp`), and MCP server (`generate_hotp`, `generate_totp`)
- Tax-Equivalent Yield Calculator: convert between a tax-free municipal bond yield and its taxable equivalent given a marginal federal tax rate plus an optional state rate, in both directions, to compare bonds on an apples-to-apples after-tax basis. Pure formula over the numbers you supply, with no live bond pricing or jurisdiction-specific tax tables. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/tax-equivalent-yield-calculator`), Node SDK (`calculateTaxEquivalentYield`), Python SDK (`calculate_tax_equivalent_yield`), and MCP server (`calculate_tax_equivalent_yield`)
- Node SDK bumped to v0.70.0, Python SDK to v0.70.0, MCP server to v0.70.0 (274 tools total, up from 272)
September 24, 2026New Tools
v0.69.0 / v0.69.0 / v0.69.0Git Commit Trailer Parser + Markdown Footnote Extractor / Validator
- Git Commit Trailer Parser: paste a full commit message and get back its trailer block (Co-authored-by, Signed-off-by, Fixes, Reviewed-by, etc.) parsed into structured key/value pairs, separated cleanly from the message body, per git's RFC 822-style trailer convention. Requires every line in the message's last paragraph to look like a trailer or an indented continuation, so a mixed paragraph is reported as having no trailers rather than guessing, unlike git's own looser 25%-of-lines heuristic. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/git-commit-trailer-parser`), Node SDK (`parseCommitTrailers`), Python SDK (`parse_commit_trailers`), and MCP server (`parse_commit_trailers`)
- Markdown Footnote Extractor / Validator: paste a Markdown document using Markdown-Extra/Pandoc-style footnotes and get back every `[^label]` reference and `[^label]:` definition extracted and cross-checked, flagging a reference with no matching definition and a definition that's never referenced. Supports multi-line definitions via 4-space or tab-indented continuation lines. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/markdown-footnote-extractor`), Node SDK (`extractMarkdownFootnotes`), Python SDK (`extract_markdown_footnotes`), and MCP server (`extract_markdown_footnotes`)
- Node SDK bumped to v0.69.0, Python SDK to v0.69.0, MCP server to v0.69.0 (272 tools total, up from 270)
September 23, 2026New Tools
v0.68.0 / v0.68.0 / v0.68.0Robots Meta Tag Builder + SVG Sprite Symbol Extractor
- Robots Meta Tag Builder: check the flag directives you want (noindex, nofollow, noarchive, nosnippet, notranslate, noimageindex, none, all, indexifembedded), add value directives like max-snippet, max-image-preview, max-video-preview, or unavailable_after, and optionally scope the tag to a single crawler by name (e.g. googlebot), to get a ready-to-paste `<meta name="robots" content="...">` tag with a plain-English note per directive. Also parses an existing tag back into its parts. The `<meta>` form of the same directive vocabulary as the shipped X-Robots-Tag Header Builder, the only option inside an HTML page's `<head>`. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/robots-meta-tag-builder`), Node SDK (`buildRobotsMetaTag`, `parseRobotsMetaTag`), Python SDK (`build_robots_meta_tag`, `parse_robots_meta_tag`), and MCP server (`build_robots_meta_tag`, `parse_robots_meta_tag`)
- SVG Sprite Symbol Extractor: paste an SVG icon sprite sheet and get back every `<symbol>` it defines: its id, viewBox, width, and height, alongside a ready-to-paste `<use href="#id"></use>` usage snippet. A symbol with no id is skipped and counted separately since it can't be referenced by a `<use>` element at all. Distinct from the shipped SVG Viewer/Editor, which analyzes, sanitizes, and optimizes a single SVG rather than enumerating a sprite sheet's individual symbols. A plain deterministic tool: no LLM calls, pure regex-based parsing over the raw markup. Shipped across every surface: browser UI, REST API (`POST /v1/tools/svg-sprite-symbol-extractor`), Node SDK (`extractSvgSpriteSymbols`), Python SDK (`extract_svg_sprite_symbols`), and MCP server (`extract_svg_sprite_symbols`)
- Node SDK bumped to v0.68.0, Python SDK to v0.68.0, MCP server to v0.68.0 (270 tools total, up from 268)
September 20, 2026New Tools
v0.67.0 / v0.67.0 / v0.67.0Regex Backreference & Named Group Visualizer + Cron Expression Generator (AI)
- Regex Backreference & Named Group Visualizer: paste a regex pattern and get back every capture group (numbered and named) it defines, every backreference it contains, and a flag on any backreference pointing at a group that doesn't exist or hasn't finished being defined yet at that point in the pattern (a forward reference or self-reference). Distinct from the shipped Regex Explainer (AI), which gives a general prose breakdown, and the ReDoS Detector, which checks backtracking risk instead of group correctness. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/regex-groups`), Node SDK (`analyzeRegexGroups`), Python SDK (`analyze_regex_groups`), and MCP server (`analyze_regex_groups`)
- Cron Expression Generator (AI): describe a schedule in plain English, like "every weekday at 9am", and get back a standard 5-field cron expression, a plain-English recap of exactly when it fires, and a field-by-field breakdown, powered by GPT-4o mini. The reverse direction of the existing Cron Parser, which explains an expression you already have rather than generating one from a description. Shipped on browser UI and REST API (`POST /v1/ai/cron-generate`), following the same Pro-gated, API-key pattern as the site's other AI generation tools (free: 3/month, Pro: 25/month)
- Node SDK bumped to v0.67.0, Python SDK to v0.67.0, MCP server to v0.67.0 (268 tools total, up from 266)
September 19, 2026New Tools
v0.66.0 / v0.66.0 / v0.66.0HTTP X-Robots-Tag Header Builder + WebSocket Frame Opcode Reference
- HTTP X-Robots-Tag Header Builder: check the flag directives you want (noindex, nofollow, noarchive, nosnippet, notranslate, noimageindex, none, all, indexifembedded), add value directives like max-snippet, max-image-preview, max-video-preview, or unavailable_after, and optionally scope the whole header to a single crawler (e.g. googlebot), to get a plain-English note per directive plus the assembled header. Also parses an existing header back into its parts. The header form of the same directive vocabulary as a robots meta tag, the only option for non-HTML resources like PDFs and images. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/x-robots-tag-builder`), Node SDK (`buildXRobotsTag`, `parseXRobotsTag`), Python SDK (`build_x_robots_tag`, `parse_x_robots_tag`), and MCP server (`build_x_robots_tag`, `parse_x_robots_tag`)
- WebSocket Frame Opcode Reference: look up what a WebSocket frame opcode (0x0 continuation, 0x1 text, 0x2 binary, 0x8 close, 0x9 ping, 0xA pong, plus reserved ranges) means per RFC 6455, or browse every defined opcode in the 4-bit opcode space at once. Distinct from the shipped WebSocket Close Code Reference, which covers close-frame status codes rather than the frame type byte itself. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/websocket-opcode-reference`), Node SDK (`lookupOpcode`, `listOpcodes`), Python SDK (`lookup_opcode`, `list_opcodes`), and MCP server (`lookup_websocket_opcode`, `list_websocket_opcodes`)
- Node SDK bumped to v0.66.0, Python SDK to v0.66.0, MCP server to v0.66.0 (266 tools total, up from 264)
September 18, 2026New Tools
v0.65.0 / v0.65.0 / v0.65.0Strict-Transport-Security (HSTS) Header Builder + Markdown Table Column Aligner
- Strict-Transport-Security (HSTS) Header Builder: build a Strict-Transport-Security response header from a max-age in seconds plus the optional includeSubDomains and preload directives, with combinations that won't qualify for a browser's HSTS preload list flagged. Also parses an existing header back into its parts. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/hsts-header-builder`), Node SDK (`buildHsts`, `parseHsts`), Python SDK (`build_hsts`, `parse_hsts`), and MCP server (`build_hsts_header`, `parse_hsts_header`)
- Markdown Table Column Aligner: reformat every GitHub-Flavored-Markdown pipe table in a document so each column's pipes line up visually, preserving left/center/right alignment markers and rebuilding the delimiter row to match, while leaving tables inside fenced code blocks untouched. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/markdown-table-aligner`), Node SDK (`alignMarkdownTables`), Python SDK (`align_markdown_tables`), and MCP server (`align_markdown_tables`)
- Node SDK bumped to v0.65.0, Python SDK to v0.65.0, MCP server to v0.65.0 (264 tools total, up from 262)
September 17, 2026New Tools
v0.64.0 / v0.64.0 / v0.64.0Loan Refinance Break-Even Calculator + Multipart/form-data Body Parser
- Loan Refinance Break-Even Calculator: compare a current loan's remaining payments against a new refinanced loan's rate, term, and closing costs, and compute the monthly savings plus the break-even month where cumulative savings recoup the closing costs. Pure time-value-of-money arithmetic, no live rate feed. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/loan-refinance-calculator`), Node SDK (`calculateLoanRefinance`), Python SDK (`calculate_loan_refinance`), and MCP server (`calculate_loan_refinance`)
- Multipart/form-data Body Parser: paste a raw multipart/form-data request body plus its boundary (or a full Content-Type header to extract the boundary from) and get back each part's field name, filename, headers, and decoded value, per RFC 7578. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/multipart-form-parser`), Node SDK (`extractMultipartBoundary`, `parseMultipartFormData`), Python SDK (`extract_multipart_boundary`, `parse_multipart_form_data`), and MCP server (`parse_multipart_form_data`)
- Node SDK bumped to v0.64.0, Python SDK to v0.64.0, MCP server to v0.64.0 (262 tools total, up from 260)
September 16, 2026New Tools
v0.63.0 / v0.63.0 / v0.63.0Debt Snowball vs Avalanche Payoff Planner + package.json Dependency Range Auditor
- Debt Snowball vs Avalanche Payoff Planner: enter a list of debts (balance, interest rate, minimum payment) and a fixed extra monthly payment, and see both the snowball (smallest balance first) and avalanche (highest rate first) payoff strategies simulated side by side, month by month, with months-to-payoff, total interest, and a full payoff schedule for each. Once a debt is paid off, its minimum payment cascades into the extra amount going toward the next targeted debt, keeping the total monthly budget fixed. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/debt-payoff-planner`), Node SDK (`planDebtPayoff`), Python SDK (`plan_debt_payoff`), and MCP server (`plan_debt_payoff`)
- package.json Dependency Range Auditor: paste a package.json and get every dependency in dependencies, devDependencies, peerDependencies, and optionalDependencies classified by version range type (exact, caret, tilde, comparator range, wildcard, npm tag, git, URL, file, or link), with wildcard ranges, the non-reproducible "latest" tag, git/URL/file/link dependencies, and packages listed with conflicting or redundant ranges across sections all flagged. A natural sibling to the shipped package.json exports field validator. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/package-json-dependency-auditor`), Node SDK (`auditDependencyRanges`), Python SDK (`audit_dependency_ranges`), and MCP server (`audit_dependency_ranges`)
- Node SDK bumped to v0.63.0, Python SDK to v0.63.0, MCP server to v0.63.0 (260 tools total, up from 258)
September 15, 2026New Tools
v0.62.0 / v0.62.0 / v0.62.0HTTP Priority Header (RFC 9218) Parser / Builder + WebSocket Close Code Reference / Lookup
- HTTP Priority Header (RFC 9218) Parser / Builder: build a `Priority` request/response-header value from an urgency (0-7, default 3) and/or an incremental flag, the header that replaced HTTP/2's old stream-dependency priority tree, or paste an existing header to see its urgency and incremental fields with spec defaults applied to whichever one is missing. A natural sibling to the site's Prefer header builder. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/priority-header`), Node SDK (`buildPriorityHeader`, `parsePriorityHeader`), Python SDK (`build_priority_header`, `parse_priority_header`), and MCP server (`build_priority_header`, `parse_priority_header`)
- WebSocket Close Code Reference / Lookup: enter a WebSocket close code and get back its name, meaning, and which reserved range it falls in, covering all 16 well-known codes from RFC 6455 (1000-1015) plus the library/framework (3000-3999) and private-use (4000-4999) IANA reserved ranges, with codes like 1005 and 1006 flagged as internal-only values that must never appear on the wire. A static reference in the same spirit as the site's HTTP Status Codes tool. Shipped across every surface: browser UI, REST API (`POST /v1/tools/websocket-close-codes`), Node SDK (`lookupCloseCode`, `listCloseCodes`), Python SDK (`lookup_close_code`, `list_close_codes`), and MCP server (`lookup_close_code`, `list_close_codes`)
- Node SDK bumped to v0.62.0, Python SDK to v0.62.0, MCP server to v0.62.0 (258 tools total, up from 256)
September 14, 2026New Tools
v0.61.0 / v0.61.0 / v0.61.0Extra-Payment Mortgage Payoff Calculator + Shell Command Explainer (AI)
- Extra-Payment Mortgage Payoff Calculator: enter a loan's principal, annual rate, and term alongside a fixed extra amount paid every month, and see the accelerated payoff time and total interest saved compared to the baseline schedule side by side. Reuses the same amortizing-loan formula as the Loan / Mortgage Calculator, then simulates the extra principal payment month by month until the balance hits zero. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/extra-payment-mortgage-calculator`), Node SDK (`calculateExtraPaymentPayoff`), Python SDK (`calculate_extra_payment_payoff`), and MCP server (`calculate_extra_payment_payoff`)
- Shell Command Explainer (AI): paste a shell or bash command and get a plain-English summary plus a flag-by-flag, argument-by-argument breakdown, with a warning surfaced when the command is destructive, irreversible, or needs elevated privileges. The reverse direction of the existing CLI Builder tool. Powered by GPT-4o mini (Free: 3/month, Pro: 25/month). Shipped on the browser UI and REST API (`POST /v1/ai/shell-explain`) only, matching the pattern of the site's other AI-generation tools
- Node SDK bumped to v0.61.0, Python SDK to v0.61.0, MCP server to v0.61.0 (256 tools total, up from 254)
September 13, 2026New Tools
v0.60.0 / v0.60.0 / v0.60.0HTTP Retry-After Header Builder + Freelancer Hourly Rate Calculator
- HTTP Retry-After Header Builder: build a `Retry-After` response header from either a plain delay in seconds or a target date/time, the header a 429 Too Many Requests or 503 Service Unavailable response uses to tell a client when it's safe to retry. A date is automatically formatted as the HTTP-date syntax the header actually requires, and an existing header value can be parsed back into the same shape. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/retry-after-builder`), Node SDK (`buildRetryAfterHeader`, `parseRetryAfterHeader`), Python SDK (`build_retry_after_header`, `parse_retry_after_header`), and MCP server (`build_retry_after_header`, `parse_retry_after_header`)
- Freelancer Hourly Rate Calculator: enter a target annual take-home income, billable hours per week, and weeks actually worked per year to get the minimum hourly rate needed to hit it, with optional annual overhead and a profit margin/buffer folded into the required revenue before it's divided across billable hours. Also returns a daily rate and a monthly revenue target. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/freelancer-rate-calculator`), Node SDK (`calculateFreelancerRate`), Python SDK (`calculate_freelancer_rate`), and MCP server (`calculate_freelancer_rate`)
- Node SDK bumped to v0.60.0, Python SDK to v0.60.0, MCP server to v0.60.0 (254 tools total, up from 252)
September 12, 2026New Tools
v0.59.0 / v0.59.0 / v0.59.0HTTP Prefer Header (RFC 7240) Parser / Builder + NDJSON Record Deduplicator
- HTTP Prefer Header (RFC 7240) Parser / Builder: add one or more preferences like `respond-async`, `wait`, or `return` with an optional value and parameters to get a ready-to-send Prefer header value, with any non-token value quoted automatically, or paste an existing Prefer header to see its preferences, values, and parameters broken out individually. A natural sibling to the site's other request-header builders (Accept-Language, Vary, Client Hints). A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/prefer-header`), Node SDK (`buildPreferHeader`, `parsePreferHeader`), Python SDK (`build_prefer_header`, `parse_prefer_header`), and MCP server (`build_prefer_header`, `parse_prefer_header`)
- NDJSON Record Deduplicator: paste NDJSON (newline-delimited JSON) and remove duplicate records in one pass, either by exact structural equality across the whole record or by a chosen key field, keeping the first occurrence of each duplicate and reporting exactly which line numbers were dropped. Distinct from the existing NDJSON Formatter / Validator, which validates and reformats but doesn't dedupe. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/ndjson-dedup`), Node SDK (`dedupeNdjson`), Python SDK (`dedupe_ndjson`), and MCP server (`dedupe_ndjson`)
- Node SDK bumped to v0.59.0, Python SDK to v0.59.0, MCP server to v0.59.0 (252 tools total, up from 250)
September 11, 2026New Tools
v0.57.0 / v0.57.0 / v0.58.0JSON Merge Patch (RFC 7396) Applier + CSP Nonce Generator
- JSON Merge Patch (RFC 7396) Applier: paste a target JSON document and an RFC 7396 merge patch document to see the merged result, with a null-valued patch key removing the matching target key and a non-object patch replacing the target wholesale, exactly per spec. Distinct from the existing JSON Pointer Resolver, which only resolves a path rather than applying a patch document. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/json-merge-patch`), Node SDK (`mergeJsonPatch`, `applyJsonMergePatch`), Python SDK (`merge_json_patch`, `apply_json_merge_patch`), and MCP server (`apply_json_merge_patch`)
- CSP Nonce Generator: generate a cryptographically random, base64-encoded nonce for a Content-Security-Policy script-src/style-src directive, with a ready-to-paste header directive and matching `<script nonce="...">`/`<style nonce="...">` tag. Fills the gap the existing CSP Header Builder leaves open, since that tool only supports a nonce placeholder and notes the real value must be generated per request. Shipped across every surface: browser UI, REST API (`POST /v1/tools/csp-nonce-generator`), Node SDK (`generateCspNonce`), Python SDK (`generate_csp_nonce`), and MCP server (`generate_csp_nonce`)
- Node SDK bumped to v0.57.0, Python SDK to v0.57.0, MCP server to v0.58.0 (250 tools total, up from 248)
September 10, 2026New Tools
v0.56.0 / v0.56.0 / v0.57.0Accept-Language Header Parser / Builder + Blended Loan Interest Rate Calculator
- Accept-Language Header Parser / Builder: add locales with an optional quality weight to get a ready-to-send Accept-Language header, sorted by descending preference per RFC 9110 with the default weight of 1 omitted the way real clients write it, or paste an existing header to see its ranked locale/quality breakdown instead. A natural sibling to the site's other request/response header builders (Vary, Permissions-Policy, Client Hints). Shipped across every surface: browser UI, REST API (`POST /v1/tools/accept-language-parser`), Node SDK (`buildAcceptLanguage`, `parseAcceptLanguage`), Python SDK (`build_accept_language`, `parse_accept_language`), and MCP server (`build_accept_language`, `parse_accept_language`)
- Blended Loan Interest Rate Calculator: enter a list of loan balances and their interest rates (e.g. several credit cards or student loans considered for consolidation) to get the single balance-weighted average rate across all of them, plus total balance and combined monthly and annual interest accrual. Distinct from the existing single-loan calculators (Loan/Mortgage, EMI), which size one loan's payment rather than answering what you're paying across several loans combined. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/blended-loan-rate-calculator`), Node SDK (`calculateBlendedLoanRate`), Python SDK (`calculate_blended_loan_rate`), and MCP server (`calculate_blended_loan_rate`)
- Node SDK bumped to v0.56.0, Python SDK to v0.56.0, MCP server to v0.57.0 (248 tools total, up from 246)
September 9, 2026New Tools
v0.55.0 / v0.55.0 / v0.56.0Debt-to-Income Ratio Calculator + CSV Column Type Inferencer
- Debt-to-Income Ratio Calculator: enter one or more monthly debts (rent or mortgage, car loan, student loans, credit cards) and your gross monthly income to see each debt's share of income plus the combined back-end DTI percentage and a qualitative tier (healthy, moderate, high, severe) matching common mortgage-lending guidance. Distinct from the existing loan/mortgage and EMI calculators, which size a single loan payment rather than answering whether you can afford more debt at all. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/debt-to-income-ratio-calculator`), Node SDK (`calculateDebtToIncomeRatio`), Python SDK (`calculate_debt_to_income_ratio`), and MCP server (`calculate_debt_to_income_ratio`)
- CSV Column Type Inferencer: paste or upload CSV/TSV data and see each column's inferred data type (integer, float, boolean, ISO 8601 date, or string), whether it's nullable, and how many empty cells it has, useful before writing a schema or an import script by hand. Distinct from the existing CSV Viewer, which displays and edits CSV but doesn't analyze column types. Shipped across every surface: browser UI, REST API (`POST /v1/tools/csv-column-type-inferencer`), Node SDK (`inferCsvColumnTypes`), Python SDK (`infer_csv_column_types`), and MCP server (`infer_csv_column_types`)
- Node SDK bumped to v0.55.0, Python SDK to v0.55.0, MCP server to v0.56.0 (246 tools total, up from 244)
September 8, 2026New Tools
v0.54.0 / v0.54.0 / v0.55.0package.json "exports" Field Validator + X-Content-Type-Options Header Builder
- package.json "exports" Field Validator: paste a package.json and get its exports field checked against the structural rules Node's module resolver actually enforces, a `types` condition not listed first, a missing `default` fallback, a `require` condition pointing at a `.mjs` file or an `import` condition pointing at a `.cjs` file, subpath and condition keys mixed in the same object, and a pattern key like `./features/*` whose target has no `*` to receive the matched text. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/package-json-exports-validator`), Node SDK (`validatePackageExports`), Python SDK (`validate_package_exports`), and MCP server (`validate_package_exports`)
- X-Content-Type-Options / MIME-Sniffing Header Builder: the header only has one valid value (`nosniff`), so this tool explains the MIME-sniffing attack it prevents, validates a pasted header value against typos, and rates how risky a given Content-Type is to serve without it. A natural sibling to the site's other security header builders (CSP, HSTS, Referrer-Policy, Permissions-Policy). Shipped across every surface: browser UI, REST API (`POST /v1/tools/x-content-type-options-builder`), Node SDK (`buildXContentTypeOptionsHeader`), Python SDK (`build_x_content_type_options_header`), and MCP server (`build_x_content_type_options_header`)
- Node SDK bumped to v0.54.0, Python SDK to v0.54.0, MCP server to v0.55.0 (244 tools total, up from 242)
September 7, 2026New Tools
v0.53.0 / v0.53.0 / v0.54.0Credit Utilization Ratio Calculator + SRT/VTT Caption Reading Speed (WPM) Checker
- Credit Utilization Ratio Calculator: enter one or more credit accounts, each with a balance and a limit, to see per-account and overall utilization percentage plus a qualitative tier (excellent, good, fair, poor) matching common credit-scoring guidance. Overall utilization is total balance over total limit, not an average of the per-account percentages, matching how credit scoring models actually weigh it. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/credit-utilization-calculator`), Node SDK (`calculateCreditUtilization`), Python SDK (`calculate_credit_utilization`), and MCP server (`calculate_credit_utilization`)
- SRT/VTT Caption Reading Speed (WPM) Checker: paste or upload an SRT or WebVTT file to see the words-per-minute and characters-per-second reading speed each cue actually demands, with cues flagged above adjustable thresholds (default ~180 WPM / ~20 CPS, common broadcast accessibility guidelines). Distinct from the existing SRT/WebVTT Subtitle Converter, which converts formats and shifts timing but doesn't evaluate reading speed. Shipped across every surface: browser UI, REST API (`POST /v1/tools/caption-reading-speed-checker`), Node SDK (`analyzeCaptionSpeed`), Python SDK (`analyze_caption_speed`), and MCP server (`analyze_caption_speed`)
- Node SDK bumped to v0.53.0, Python SDK to v0.53.0, MCP server to v0.54.0 (242 tools total, up from 240)
September 6, 2026New Tools
v0.52.0 / v0.52.0 / v0.53.0.editorconfig Validator + tsconfig.json Validator
- .editorconfig Validator: paste an .editorconfig file and see it parsed into its root flag and glob sections, with every syntax mistake flagged by line number, an invalid indent_style value, a root declaration placed after a section instead of in the preamble, a malformed line, or a duplicate property. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/editorconfig-validator`), Node SDK (`validateEditorConfig`), Python SDK (`validate_editorconfig`), and MCP server (`validate_editorconfig`)
- tsconfig.json Validator: paste a tsconfig.json file (JSONC comments and trailing commas are stripped first, the same leniency tsc itself applies) and get back every structural mistake checked against a curated list of known compilerOptions, a type mismatch like `strict` given a string instead of a boolean, an enum option like `target` given a value outside its known set, or a malformed `include`/`exclude`/`references`/`paths`. Distinct from a plain JSON formatter, since it checks TypeScript-specific option types and values, not just JSON syntax. Shipped across every surface: browser UI, REST API (`POST /v1/tools/tsconfig-validator`), Node SDK (`validateTsconfig`), Python SDK (`validate_tsconfig`), and MCP server (`validate_tsconfig`)
- Node SDK bumped to v0.52.0, Python SDK to v0.52.0, MCP server to v0.53.0 (240 tools total, up from 238)
September 5, 2026New Tools
v0.51.0 / v0.51.0 / v0.52.0GPA Calculator + PR Description Generator (AI)
- Weighted Grade / GPA Calculator: enter each course's letter grade and credit hours to get a weighted GPA back, using the standard 4.0-capped scale where A+ counts the same as A. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/gpa-calculator`), Node SDK (`calculateGpa`), Python SDK (`calculate_gpa`), and MCP server (`calculate_gpa`)
- PR Description Generator (AI): paste a git diff and get back a structured pull request description, a title, a bulleted summary of what changed, and a test plan checklist, powered by GPT-4o mini. Distinct from the existing Commit Message Generator, which writes a short Conventional Commits message for a single commit rather than the longer reviewer-facing PR body. Shipped on the browser and REST API (`POST /v1/ai/pr-description`) only, the same pattern as the site's other AI generation tools: no Node/Python SDK or MCP surface. Free plan: 3 generations/month, Pro: 25/month
- Node SDK bumped to v0.51.0, Python SDK to v0.51.0, MCP server to v0.52.0 (238 tools total, up from 236)
August 26, 2026New Tools
v0.50.0 / v0.50.0 / v0.51.0Moving Average Calculator + Markdown Task List Progress Calculator
- Simple Moving Average / Exponential Moving Average Calculator: paste a series of numbers and a window size to get SMA and EMA at every point, plus the latest value of each, a foundational technical-indicator building block for price or volume data with no live market-data feed involved. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/moving-average-calculator`), Node SDK (`calculateMovingAverages`), Python SDK (`calculate_moving_averages`), and MCP server (`calculate_moving_averages`)
- Markdown Task List Progress Calculator: paste a Markdown document and get completion stats for every `- [ ]`/`- [x]` checkbox, overall and broken down per section, where a section is the nearest preceding heading of any level. Distinct from the existing Table of Contents Generator and Front Matter Parser, neither of which touches task lists. Shipped across every surface: browser UI, REST API (`POST /v1/tools/markdown-task-list-progress`), Node SDK (`calculateMarkdownTaskListProgress`), Python SDK (`calculate_markdown_task_list_progress`), and MCP server (`calculate_markdown_task_list_progress`)
- Node SDK bumped to v0.50.0, Python SDK to v0.50.0, MCP server to v0.51.0 (236 tools total, up from 234)
August 25, 2026New Tools
v0.49.0 / v0.49.0 / v0.50.0Cash Rounding Calculator + Test Case Generator (AI)
- Cash Rounding (Swedish Rounding) Calculator: enter a cashless total and the smallest payable cash denomination to get back the rounded total and the exact adjustment, with round-nearest, round-up, and round-down modes. Distinct from the existing Cash Denomination / Till Calculator, which breaks an already-known total into bill/coin counts rather than deciding what it should round to. A plain deterministic tool: no LLM calls. Shipped across every surface: browser UI, REST API (`POST /v1/tools/cash-rounding-calculator`), Node SDK (`calculateCashRounding`), Python SDK (`calculate_cash_rounding`), and MCP server (`calculate_cash_rounding`)
- Test Case Generator (AI): paste a function and get back a set of proposed unit test cases, a description, concrete input, expected output, and whether each covers a happy path, edge case, or error case, powered by GPT-4o mini. Returns structured test case data rather than runnable test code in a specific framework. Shipped on the browser and REST API (`POST /v1/ai/testcase`) only, the same pattern as the site's other AI generation tools: no Node/Python SDK or MCP surface. Free plan: 3 generations/month, Pro: 25/month
- Node SDK bumped to v0.49.0, Python SDK to v0.49.0, MCP server to v0.50.0 (234 tools total, up from 232)
August 24, 2026New Tools
v0.48.0 / v0.48.0 / v0.49.0YAML Anchor & Alias Resolver + Sitemap Index Splitter
- YAML Anchor & Alias Resolver: paste YAML using `&anchor` definitions, `*alias` references, or `<<:` merge keys to get back the fully expanded document with every reference resolved inline, plus a table of which anchors were referenced and which were left unused. Distinct from the existing YAML Formatter/Validator, which only checks syntax and never expands anchors
- Sitemap Index Splitter: paste a URL list that exceeds the sitemaps.org 50,000-URL cap along with a base URL to generate a set of compliant sitemap-N.xml files plus the sitemap index file that references them. The generator-side sibling of the existing Sitemap.xml Validator, which only flags the cap as an error
- Both are plain deterministic tools, not AI-powered: no LLM calls, no external API keys needed. Shipped across every surface: browser UI, REST API (`POST /v1/tools/yaml-anchor-resolver`, `POST /v1/tools/sitemap-index-splitter`), Node SDK (`resolveYamlAnchors`, `splitSitemap`), Python SDK (`resolve_yaml_anchors`, `split_sitemap`), and MCP server (`resolve_yaml_anchors`, `split_sitemap`)
- Node SDK bumped to v0.48.0, Python SDK to v0.48.0, MCP server to v0.49.0 (232 tools total, up from 230)
August 23, 2026New Tools
v0.47.0 / v0.47.0 / v0.48.0Content-Disposition Header Builder + GIF Frame & Loop Inspector
- Content-Disposition Header Builder: pick inline or attachment and enter a filename to get the assembled `Content-Disposition` header back, with an RFC 5987 `filename*` extended parameter automatically added alongside an ASCII-sanitized fallback whenever the filename has non-ASCII characters. Also parses an existing header you paste in into its type, decoded filename, and raw parameters
- GIF Frame & Loop Inspector: drop in a GIF file to see its frame count, total animation duration, loop count, and global color table info, read directly from the file's GIF87a/89a block structure (image descriptors, graphic control extensions, the NETSCAPE2.0 application extension) without decoding any LZW-compressed pixel data
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/content-disposition-builder`, `POST /v1/tools/gif-frame-inspector`), Node SDK (`buildContentDisposition`, `parseContentDisposition`, `inspectGifFrames`), Python SDK (`build_content_disposition`, `parse_content_disposition`, `inspect_gif_frames`), and MCP server (`build_content_disposition`, `parse_content_disposition`, `inspect_gif_frames`)
- Node SDK bumped to v0.47.0, Python SDK to v0.47.0, MCP server to v0.48.0 (230 tools total, up from 228)
August 21, 2026New Tools
v0.46.0 / v0.46.0 / v0.47.0Depreciation Schedule Calculator + Keyword Density Analyzer
- Depreciation Schedule Calculator: enter an asset's cost, salvage value, and useful life in years to get a full year-by-year depreciation schedule back, using straight-line, double-declining-balance, or sum-of-years-digits. Double-declining-balance clamps at salvage value rather than switching to straight-line partway through
- Keyword Density Analyzer: paste in an article or page copy and see keyword and phrase frequency plus density (occurrences per 100 words), with single-word mode filtering common English stopwords and 2/3-word phrase modes keeping them since useful phrases often contain them
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/depreciation-schedule-calculator`, `POST /v1/tools/keyword-density-analyzer`), Node SDK (`calculateDepreciationSchedule`, `analyzeKeywordDensity`), Python SDK (`calculate_depreciation_schedule`, `analyze_keyword_density`), and MCP server (`calculate_depreciation_schedule`, `analyze_keyword_density`)
- Node SDK bumped to v0.46.0, Python SDK to v0.46.0, MCP server to v0.47.0 (228 tools total, up from 226)
August 20, 2026New Tools
v0.45.0 / v0.45.0 / v0.46.0Certificate of Deposit (CD) Maturity Calculator + Client Hints Header Builder
- Certificate of Deposit (CD) Maturity Calculator: enter a principal, nominal annual rate, term in months, and compounding frequency to get the maturity value, total interest, and effective annual yield (APY) back, plus the net payout if you cash out early with an optional penalty expressed as months of forfeited interest
- Client Hints (Sec-CH-*) Header Builder: pick which client hints (browser platform, dark-mode preference, viewport width, device memory, and more) your server wants the browser to send and get the assembled `Accept-CH` header back, with a plain-English note on what each hint reveals; mark a subset critical to also generate `Critical-CH`. Also decodes raw Client Hints request headers you paste in into readable JSON
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/cd-maturity-calculator`, `POST /v1/tools/client-hints-header-builder`), Node SDK (`calculateCdMaturity`, `buildClientHints`, `parseClientHintHeaders`), Python SDK (`calculate_cd_maturity`, `build_client_hints`, `parse_client_hint_headers`), and MCP server (`calculate_cd_maturity`, `build_client_hints`, `parse_client_hint_headers`)
- Node SDK bumped to v0.45.0, Python SDK to v0.45.0, MCP server to v0.46.0 (226 tools total, up from 224)
August 19, 2026New Tools
v0.44.0 / v0.44.0 / v0.45.0Permissions-Policy Header Builder + Payback Period Calculator
- Permissions-Policy Header Builder: pick which browser features (camera, microphone, geolocation, fullscreen, and more) your page allows and their allowlist (blocked, self only, all origins, or specific https:// origins) and get the assembled `Permissions-Policy` header back, with a plain-English note on what each directive restricts. Also parses an existing header you paste in to explain it the same way
- Payback Period Calculator: enter an initial investment and a series of periodic cash flows to see the fractional payback period (the point cumulative cash flow turns positive), plus a discounted payback period when you supply a discount rate, which discounts each cash flow before accumulating it
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/permissions-policy-builder`, `POST /v1/tools/payback-period-calculator`), Node SDK (`buildPermissionsPolicy`, `parsePermissionsPolicy`, `calculatePaybackPeriod`), Python SDK (`build_permissions_policy`, `parse_permissions_policy`, `calculate_payback_period`), and MCP server (`build_permissions_policy`, `parse_permissions_policy`, `calculate_payback_period`)
- Node SDK bumped to v0.44.0, Python SDK to v0.44.0, MCP server to v0.45.0 (224 tools total, up from 222)
August 19, 2026New Tools
v0.43.0 / v0.43.0 / v0.44.0Tool-Definition Schema Linter (AI/Agent) + Web App Manifest Validator
- Tool-Definition Schema Linter: paste an LLM tool/function-calling definition in OpenAI, Anthropic, or MCP format (a single definition, an array of them, or a `{tools:[...]}` wrapper) and get flagged the mistakes that measurably hurt model accuracy: a missing or vague description, an invalid name, a non-object top-level parameter schema, properties missing descriptions, an empty enum, a `required` entry with no matching property, or duplicate tool names
- Web App Manifest Validator: paste a manifest.json and see exactly which installability requirements it's missing before a browser or app store rejects it, checking name/short_name, start_url, display mode, and icon coverage (including a 512x512 icon and a maskable icon), with softer recommendations flagged as warnings
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/tool-schema-linter`, `POST /v1/tools/manifest-validator`), Node SDK (`lintToolSchema`, `validateManifest`), Python SDK (`lint_tool_schema`, `validate_manifest`), and MCP server (`lint_tool_schema`, `validate_manifest`)
- Node SDK bumped to v0.43.0, Python SDK to v0.43.0, MCP server to v0.44.0 (222 tools total, up from 220)
August 18, 2026New Tools
v0.42.0 / v0.42.0 / v0.43.0HTTP Vary Header Builder + Code Docstring / Comment Generator (AI)
- HTTP Vary Header Builder: pick which request headers a response varies by (Accept-Encoding, Accept-Language, Authorization, Cookie, Origin, and more, or a custom header name) and get the assembled `Vary` header value back, with a plain-English note on the caching implication of each header and a warning when `*` is combined with other names. Also parses an existing Vary header you paste in to explain it the same way
- Code Docstring / Comment Generator (AI): paste a function or class and get a documentation comment written in the idiomatic style for its language (JSDoc for JavaScript/TypeScript, a Google-style docstring for Python, Javadoc, rustdoc, and similar), plus the original code with the comment inserted above it. Powered by GPT-4o mini, 3 free generations/month, 25/month on Pro
- HTTP Vary Header Builder shipped across every surface: browser UI, REST API (`POST /v1/tools/vary-header-builder`), Node SDK (`buildVaryHeader`, `parseVaryHeader`), Python SDK (`build_vary_header`, `parse_vary_header`), and MCP server (`build_vary_header`, `parse_vary_header`). Code Docstring Generator shipped as browser UI + REST API only (`POST /v1/ai/docstring`), matching the site's other AI-generation tools
- Node SDK bumped to v0.42.0, Python SDK to v0.42.0, MCP server to v0.43.0 (220 tools total, up from 218)
August 12, 2026New Tools
v0.41.0 / v0.41.0 / v0.42.0HTTP Range Request Header Builder + CSV Duplicate Row Finder
- HTTP Range Request Header Builder: build a `Range` request header from one or more byte ranges (explicit start-end pairs, open-ended ranges, or a suffix "last N bytes" range), with byte-length and percentage-of-file resolution when you supply a total size, plus overlap warnings. Also parses a server's `Content-Range` response so you can verify what it actually sent back for a resumable download or partial fetch
- CSV Duplicate Row Finder: paste CSV or TSV data and find duplicate rows, either exact full-row duplicates or duplicates by a chosen subset of key columns, with each duplicate group's occurrence count and exact data row numbers so cleanup before a database or spreadsheet import is a matter of jumping to those rows
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/http-range-header-builder`, `POST /v1/tools/csv-duplicate-row-finder`), Node SDK (`buildRangeHeader`, `parseContentRangeHeader`, `findCsvDuplicateRows`), Python SDK (`build_range_header`, `parse_content_range_header`, `find_csv_duplicate_rows`), and MCP server (`build_range_header`, `parse_content_range_header`, `find_csv_duplicate_rows`)
- Node SDK bumped to v0.41.0, Python SDK to v0.41.0, MCP server to v0.42.0 (218 tools total, up from 216)
August 11, 2026New Tools
v0.40.0 / v0.40.0 / v0.41.0Cash Denomination / Till Calculator + HTTP ETag Generator
- Cash Denomination / Till Calculator: enter a cash total to see the minimum count of bills and coins needed from a set of denominations (USD, EUR, and GBP presets, or a custom list), or switch to change mode to enter an amount due and amount tendered and get the change owed broken down the same way. Uses the standard greedy largest-first approach, optimal for canonical currency denomination sets
- HTTP ETag Generator: paste the exact content you'd serve and get back a properly quoted ETag header value, digested with MD5, SHA-1, or SHA-256, with an optional weak-validator (W/) prefix. A content-hash generator, not a reimplementation of any particular server's own ETag algorithm, since many servers key their ETag off file size and modification time instead
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/cash-till-calculator`, `POST /v1/tools/etag-generator`), Node SDK (`calculateCashBreakdown`, `calculateChange`, `generateEtag`), Python SDK (`calculate_cash_breakdown`, `calculate_change`, `generate_etag`), and MCP server (`calculate_cash_breakdown`, `calculate_change`, `generate_etag`)
- Node SDK bumped to v0.40.0, Python SDK to v0.40.0, MCP server to v0.41.0 (216 tools total, up from 214)
August 10, 2026New Tools
v0.39.0 / v0.39.0 / v0.40.0SRI Hash Generator + Email Header Analyzer
- SRI Hash Generator: paste a script or stylesheet's exact content and get back its Subresource Integrity hash for SHA-256, SHA-384, and SHA-512 all at once, each formatted as the `<algorithm>-<base64>` string an HTML `integrity` attribute expects. Content-based only: it never fetches a URL, so the input needs to be the exact bytes that will actually be served
- Email Header Analyzer: paste a raw email header block (Gmail's "Show original", a .eml file) and get back the common fields (From, To, Subject, Date, Message-ID), the full Received hop chain in order, and any SPF/DKIM/DMARC verdicts already present in an Authentication-Results header. It only reports what the headers already claim, it never performs a live DNS/SPF/DKIM check of its own
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/sri-hash-generator`, `POST /v1/tools/email-header-analyzer`), Node SDK (`generateSriAll`, `analyzeEmailHeaders`), Python SDK (`generate_sri_all`, `analyze_email_headers`), and MCP server (`generate_sri_hash`, `analyze_email_headers`)
- Node SDK bumped to v0.39.0, Python SDK to v0.39.0, MCP server to v0.40.0 (214 tools total, up from 212)
August 9, 2026New Tools
v0.38.0 / v0.38.0 / v0.39.0Unit Price Comparator + CSV to SQL INSERT Statement Generator
- Unit Price Comparator: enter two or more package options' price and quantity to find the true cost per unit for each, see the cheapest one flagged, and see exactly how much more expensive every other option is by percentage. Unit-agnostic by design: it never converts between units, so grams, litres, or item counts all work as long as every option in the comparison uses the same one
- CSV to SQL INSERT Statement Generator: paste CSV data with a header row and a table name to get ready-to-run SQL INSERT statements back, with basic per-cell type inference (integers and true/false left unquoted, empty cells become NULL, everything else single-quoted with escaping) and a choice of ANSI double-quote or MySQL backtick identifier quoting
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/unit-price-comparator`, `POST /v1/tools/csv-to-sql-insert-generator`), Node SDK (`compareUnitPrices`, `csvToSqlInsert`), Python SDK (`compare_unit_prices`, `csv_to_sql_insert`), and MCP server (`compare_unit_prices`, `csv_to_sql_insert`)
- Node SDK bumped to v0.38.0, Python SDK to v0.38.0, MCP server to v0.39.0 (212 tools total, up from 210)
August 8, 2026New Tools
v0.37.0 / v0.37.0 / v0.38.0Same-Site Cookie Attribute Checker + .gitignore Pattern Tester
- Same-Site Cookie Attribute Checker: paste a Set-Cookie header value and validate its attribute combination against browser cookie rules: SameSite=None requires Secure, the __Host-/__Secure- name prefixes each carry their own required attributes, a missing Secure attribute gets flagged, and a Max-Age past 400 days gets flagged since Chromium-based browsers silently clamp it to that cap
- .gitignore Pattern Tester: paste a .gitignore file and a list of paths to see which ones git would actually ignore, following real matching rules including negation (`!`), directory-only patterns, root-anchored patterns, and the last-match-wins precedence rule that trips people up when hand-tracing a long .gitignore
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/same-site-cookie-checker`, `POST /v1/tools/gitignore-tester`), Node SDK (`checkSetCookieHeader`, `testGitignorePatterns`), Python SDK (`check_set_cookie_header`, `test_gitignore_patterns`), and MCP server (`check_set_cookie_header`, `test_gitignore_patterns`)
- Node SDK bumped to v0.37.0, Python SDK to v0.37.0, MCP server to v0.38.0 (210 tools total, up from 208)
August 7, 2026New Tools
v0.36.0 / v0.36.0 / v0.37.0Regex Explainer (AI) + HTTP Link Header Builder
- Regex Explainer (AI): paste a regex pattern (and optional flags) to get a plain-English summary of what it matches, a token-by-token breakdown of each fragment and capture group, flag meanings, and generated example strings that would and would not match, powered by GPT-4o mini. The reverse direction of the existing NL -> Regex generator: that one writes a pattern from a description, this one explains a pattern someone else wrote
- HTTP Link Header Builder: add one or more URL/rel pairs (preload, preconnect, prefetch, canonical, next/prev pagination, and more) to get an assembled `Link` response header, including the `as`/`type`/`crossorigin` attributes preload needs, plus a plain-English explanation of each entry and warnings for common mistakes like a preload missing `as` or more than one canonical link. Runs entirely client-side and pairs with the existing Cache-Control, CSP, CORS, and Referrer-Policy header builders
- Regex Explainer shipped on browser UI and REST API (`POST /v1/ai/regex-explain`, 3/month free, 25/month Pro); Link Header Builder shipped across every surface: browser UI, REST API (`POST /v1/tools/link-header-builder`), Node SDK (`buildLinkHeader`), Python SDK (`build_link_header`), and MCP server (`build_link_header`)
- Node SDK bumped to v0.36.0, Python SDK to v0.36.0, MCP server to v0.37.0 (208 tools total, up from 206)
August 6, 2026New Tools
v0.35.0 / v0.35.0 / v0.36.0ROI Calculator + Emoji Lookup
- ROI Calculator: enter an initial and final value to get the return on investment as a plain percentage plus the profit or loss in the same units, and optionally add a holding period in years to also see the annualized (CAGR) return, the constant yearly growth rate that would take the initial value to the final value over that span. Distinct from the present-value and compound-interest calculators, which grow a sum forward at a known rate: this derives the realized rate from actual start/end values
- Emoji Lookup: search a curated table of ~280 common emoji by name, keyword, or partial shortcode and get back the exact character, its Unicode code point (like U+1F525), and its GitHub-style shortcode (like :fire:); also supports exact lookup by shortcode, character, or category. Not the full 3,000+ Unicode emoji set, just the most commonly used ones for chat, commit messages, and markdown
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/roi-calculator`, `POST /v1/tools/emoji-lookup`), Node SDK (`calculateRoi`, `searchEmoji`/`getEmojiByShortcode`/`getEmojiByChar`/`getEmojiByCategory`), Python SDK (`calculate_roi`, `search_emoji`/`get_emoji_by_shortcode`/`get_emoji_by_char`/`get_emoji_by_category`), and MCP server (`calculate_roi`, `lookup_emoji`)
- Node SDK bumped to v0.35.0, Python SDK to v0.35.0, MCP server to v0.36.0 (206 tools total, up from 204)
August 5, 2026New Tools
v0.34.0 / v0.34.0 / v0.35.0Referrer-Policy Header Builder + Markdown Table of Contents Generator
- Referrer-Policy Header Builder: pick one or more of the 8 standard Referrer-Policy tokens (forming a fallback chain) to get the assembled header value, an equivalent `<meta name="referrer">` tag, and a plain-English explanation of exactly what gets sent to a destination under each token, including whether it still leaks on an HTTPS to HTTP downgrade
- Markdown Table of Contents Generator: paste a Markdown document to get a nested bullet or numbered table of contents built from its ATX headings, with GitHub-style anchor slugs; headings inside fenced code blocks are ignored and duplicate headings get `-1`, `-2`, ... suffixes the same way GitHub disambiguates them
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/referrer-policy-builder`, `POST /v1/tools/markdown-toc-generator`), Node SDK (`buildReferrerPolicy`, `generateMarkdownToc`), Python SDK (`build_referrer_policy`, `generate_markdown_toc`), and MCP server (`build_referrer_policy`, `generate_markdown_toc`)
- Node SDK bumped to v0.34.0, Python SDK to v0.34.0, MCP server to v0.35.0 (204 tools total, up from 202)
August 4, 2026New Tools
v0.33.0 / v0.33.0 / v0.34.0Conventional Commit Message Linter + Savings Goal / Time-to-Target Calculator
- Conventional Commit Message Linter: paste a git commit message to check it against the Conventional Commits v1.0.0 spec: header format (`type(scope)!: description`), the required blank line before a body or footer, footer token formatting, and BREAKING CHANGE casing. Parses out the type, scope, breaking flag, description, body, and footers, and reports each issue as an error (a structural spec violation) or a warning (a style convention like lower-case types)
- Savings Goal / Time-to-Target Calculator: enter a goal amount, current savings, and an annual interest rate to solve for either how many months it takes to reach the goal at a fixed monthly contribution, or the monthly contribution required to reach the goal within a target number of months. Standard ordinary-annuity time-value-of-money math, also reporting total contributed, interest earned, and final balance
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/conventional-commit-linter`, `POST /v1/tools/savings-goal-calculator`), Node SDK (`lintConventionalCommit`, `calculateSavingsGoal`), Python SDK (`lint_conventional_commit`, `calculate_savings_goal`), and MCP server (`lint_conventional_commit`, `calculate_savings_goal`)
- Node SDK bumped to v0.33.0, Python SDK to v0.33.0, MCP server to v0.34.0 (202 tools total, up from 200)
August 3, 2026New Tools
v0.32.0 / v0.32.0 / v0.33.0Sentence Length Distribution Analyzer + Few-Shot Prompt Example Formatter
- Sentence Length Distribution Analyzer: paste a paragraph to get a histogram of sentence lengths in 5-word buckets, plus mean, median, standard deviation, and a coefficient-of-variation consistency label. A signal distinct from the Flesch/Gunning Fog grade-level scores the existing Readability Score Calculator reports, since two paragraphs can share the same reading grade while one alternates short and long sentences and the other stays uniform
- Few-Shot Prompt Example Formatter: add a list of input/output example pairs and format them into a numbered, XML-tagged, or delimiter-separated few-shot prompt template ready to paste into a system or user prompt. Pure string formatting with no model call involved, and pairs with an empty input or output are dropped automatically
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/sentence-length-distribution`, `POST /v1/tools/few-shot-formatter`), Node SDK (`analyzeSentenceLengths`, `formatFewShotExamples`), Python SDK (`analyze_sentence_lengths`, `format_few_shot_examples`), and MCP server (`analyze_sentence_lengths`, `format_few_shot_examples`)
- Node SDK bumped to v0.32.0, Python SDK to v0.32.0, MCP server to v0.33.0 (200 tools total, up from 198)
August 2, 2026New Tools
v0.31.0 / v0.31.0 / v0.32.0EMI Calculator + Markdown Front Matter Parser
- EMI Calculator: enter a loan principal, annual interest rate, and tenure in years or months to get the Equated Monthly Installment, total interest, and total amount payable, using the same reducing-balance amortization formula as the existing Loan / Mortgage Calculator. Results are grouped into a year-by-year principal/interest breakdown instead of a full monthly schedule, matching how EMI figures are typically presented to borrowers
- Markdown Front Matter Parser: paste a Markdown document with a `---`-delimited YAML front matter block to split it into structured fields and body content. Handles scalar strings, quoted strings, booleans, numbers, null, inline arrays, and YAML block lists; a document with no front matter is passed through unchanged rather than treated as an error
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/emi-calculator`, `POST /v1/tools/markdown-front-matter-parser`), Node SDK (`calculateEmi`, `parseFrontMatter`), Python SDK (`calculate_emi`, `parse_front_matter`), and MCP server (`calculate_emi`, `parse_markdown_front_matter`)
- Node SDK bumped to v0.31.0, Python SDK to v0.31.0, MCP server to v0.32.0 (198 tools total, up from 196)
August 1, 2026New Tools
v0.30.0 / v0.30.0 / v0.31.0JSON Schema to TypeScript Types + Password Entropy Estimator
- JSON Schema to TypeScript Types: paste a JSON Schema document to generate matching TypeScript interfaces, reading the schema's own `required`, `enum`, `const`, `oneOf`/`anyOf`/`allOf`, and `$ref` keywords directly instead of guessing types from a single example payload the way the existing JSON to TypeScript tool does. Optional vs. required fields, literal/union types, and reusable sub-schemas via local `$ref` into `$defs`/`definitions` all come through accurately
- Password Entropy Estimator: type or paste a password to see its actual entropy in bits, computed from length and the character pool it draws from (lowercase, uppercase, digits, symbols, other/Unicode), a strength label, and estimated crack time across four scenarios from a throttled login form to an offline GPU cracking cluster. Also flags repeated character runs, sequential runs, short length, and single-character-class passwords, since raw entropy alone overestimates resistance for guessable patterns. Runs entirely client-side; nothing is transmitted
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/json-schema-to-typescript`, `POST /v1/tools/password-entropy-estimator`), Node SDK (`jsonSchemaToTypescript`, `estimatePasswordEntropy`), Python SDK (`json_schema_to_typescript`, `estimate_password_entropy`), and MCP server (`json_schema_to_typescript`, `estimate_password_entropy`)
- Node SDK bumped to v0.30.0, Python SDK to v0.30.0, MCP server to v0.31.0 (196 tools total, up from 194)
July 31, 2026New Tools
v0.29.0 / v0.29.0 / v0.30.0PDF Page Dimension & Orientation Lister + IPv6 Address Compressor / Expander
- PDF Page Dimension & Orientation Lister: drop in a PDF to see every page's width and height in points and inches, its rotation, and whether it reads as portrait or landscape, plus a match against common paper sizes like Letter, Legal, or A4. It walks the actual Root -> Pages -> Kids page tree and inherits /MediaBox and /Rotate from ancestor nodes the way a real PDF viewer does, so pages that inherit their size instead of setting their own still come back correct
- IPv6 Address Compressor / Expander: paste any IPv6 address to see both its full expanded form (all 8 groups) and its RFC 5952 canonical compressed form side by side, plus its address type: loopback, link-local, unique local, multicast, IPv4-mapped, documentation, or ordinary global unicast. Understands a zone ID like %eth0 and an embedded IPv4 suffix like ::ffff:192.168.1.1
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/pdf-page-dimensions`, `POST /v1/tools/ipv6-compressor`), Node SDK (`readPdfPageDimensions`, `parseIpv6Address`), Python SDK (`read_pdf_page_dimensions`, `parse_ipv6_address`), and MCP server (`read_pdf_page_dimensions`, `parse_ipv6_address`)
- Node SDK bumped to v0.29.0, Python SDK to v0.29.0, MCP server to v0.30.0 (194 tools total, up from 192)
July 30, 2026New Tools
v0.28.0 / v0.28.0 / v0.29.0MP4/WebM Video Info + Error Message Explainer (AI)
- MP4/WebM Video Info: drop in an MP4 or WebM file to read its duration, resolution, and video/audio codec identifiers straight from the container header, the moov/mvhd/trak boxes for MP4 (ISO BMFF) or the Segment/Info/Tracks elements for WebM (EBML/Matroska). No frame is ever decoded, no sample data is demuxed, and nothing is transcoded, so a two-hour 4K file returns just as fast as a two-second clip
- Error Message Explainer (AI): paste a compiler, linker, or package-manager error, a TypeScript type error, an npm ERESOLVE conflict, a Rust borrow-checker complaint, and get back a plain-English explanation of what it means, the usual root causes, and a concrete next step, powered by GPT-4o mini. This reads a build-time diagnostic rather than a runtime exception, which is what the existing Stack Trace Summarizer covers instead
- MP4/WebM Video Info shipped across every surface: browser UI, REST API (`POST /v1/tools/mp4-webm-video-info`, multipart file upload), Node SDK (`readVideoInfo`), Python SDK (`read_video_info`), and MCP server (`read_video_info`)
- Error Message Explainer (AI) shipped to browser UI and REST API (`POST /v1/ai/error-message`), matching the site's other GPT-4o mini tools: no Node/Python SDK or MCP wrapper, since the generation itself has to run server-side
- Node SDK bumped to v0.28.0, Python SDK to v0.28.0, MCP server to v0.29.0 (192 tools total, up from 190)
July 29, 2026New Tools
v0.27.0 / v0.27.0 / v0.28.0Rule of 72 / Doubling Time Calculator + FLAC Metadata Reader
- Rule of 72 / Doubling Time Calculator: enter an annual rate to estimate how many years it takes an investment to double, or flip it around and enter a number of years to estimate the rate needed to double in that time. Both directions also show the exact result from the logarithmic doubling-time formula alongside the Rule of 72 shortcut, plus how far apart they are: the shortcut is closest to exact for rates roughly between 6% and 10%
- FLAC Metadata Reader: drop in a FLAC file to read its sample rate, channel count, bit depth, total samples, and duration straight from the STREAMINFO metadata block, no audio decoding required. If the file carries a VORBIS_COMMENT block, the encoder vendor string and tags like artist, title, album, date, genre, and track number come back too, with the full raw tag list for anything less common
- Files with no VORBIS_COMMENT block simply come back without tag fields rather than erroring, since tags are optional in the FLAC format; the reader expects the native "fLaC" container, not FLAC audio wrapped in an Ogg container
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/rule-of-72-calculator`, `POST /v1/tools/flac-metadata-reader`), Node SDK (`calculateRuleOf72`, `readFlacInfo`), Python SDK (`calculate_rule_of_72`, `read_flac_info`), and MCP server (`calculate_rule_of_72`, `read_flac_info`)
- Node SDK bumped to v0.27.0, Python SDK to v0.27.0, MCP server to v0.28.0 (143 tools total, up from 141)
July 28, 2026New Tools
v0.26.0 / v0.26.0 / v0.27.0JWK Thumbprint Calculator + Video Bitrate & File Size Calculator
- JWK Thumbprint Calculator: paste a single JWK and get its RFC 7638 canonical thumbprint, the fingerprint used as a key ID in OpenID Connect and to bind a token to a key in DPoP and mutual-TLS. Only the required members per key type (e.g. e / kty / n for RSA) are hashed, in lexicographic order with no whitespace, so optional fields like alg or kid never change the result
- Supports RSA, EC, oct (symmetric), and OKP (Ed25519/X25519) keys, with SHA-256, SHA-384, or SHA-512, and shows the exact canonical JSON that went into the hash. Rejects a JWKS set outright, since a thumbprint is defined per key, not per set
- Video Bitrate & File Size Calculator: enter a video bitrate, an optional audio bitrate, and a duration to get the resulting file size, or flip it around and enter a target file size to get the bitrate required to hit it. Results show both decimal MB and binary MiB, since file managers display the latter
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/jwk-thumbprint`, `POST /v1/tools/video-bitrate-calculator`), Node SDK (`computeJwkThumbprint`, `calculateVideoBitrate`), Python SDK (`compute_jwk_thumbprint`, `calculate_video_bitrate`), and MCP server (`compute_jwk_thumbprint`, `calculate_video_bitrate`)
- Node SDK bumped to v0.26.0, Python SDK to v0.26.0, MCP server to v0.27.0 (141 tools total, up from 139)
July 27, 2026New Tools
v0.25.0 / v0.25.0 / v0.26.0Break-Even Point Calculator + Levenshtein / String Distance Calculator
- Break-Even Point Calculator: enter fixed costs, the variable cost per unit, and the price per unit to get the break-even unit count and revenue, the contribution margin per unit and as a percent of price, and, if you supply a sales volume, the resulting profit or loss and margin of safety above break-even
- A price that does not exceed the variable cost per unit is rejected outright, since no sales volume would ever reach break-even in that case; the break-even unit count is also shown rounded up to the next whole unit alongside the exact figure, since revenue math and "how many do I need to sell" are two different questions
- Levenshtein / String Distance Calculator: enter two strings to get the Levenshtein edit distance between them plus a normalized 0-1 similarity ratio. Switch to Damerau mode to also count one adjacent-character transposition (like "ab" to "ba") as a single edit instead of two substitutions
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/break-even-calculator`, `POST /v1/tools/string-distance`), Node SDK (`calculateBreakEven`, `calculateStringDistance`), Python SDK (`calculate_break_even`, `calculate_string_distance`), and MCP server (`calculate_break_even`, `calculate_string_distance`)
- Node SDK bumped to v0.25.0, Python SDK to v0.25.0, MCP server to v0.26.0 (139 tools total, up from 137)
July 26, 2026New Tools
v0.24.0 / v0.24.0 / v0.25.0Discount Calculator + SRT / WebVTT Subtitle Converter
- Discount Calculator: enter an original price plus any one of the discount percent, the amount off, or the final sale price, and the other two are solved for you. The second mode stacks discounts applied in sequence and shows each step, so you can see that 20% off followed by an extra 10% off is a 28% discount rather than 30%, because the second cut comes off the already-reduced price
- SRT / WebVTT Subtitle Converter: paste or upload a subtitle file and convert between SubRip and WebVTT, with the source format detected automatically from the WEBVTT header or the timestamp separator. An optional time shift moves every cue by a set number of seconds to resync it against a different cut, clamping anything pushed below zero to 00:00:00
- Converting WebVTT to SRT drops cue identifiers, cue settings, and NOTE / STYLE / REGION blocks, since SubRip has no equivalent for any of them. Each drop, along with overlapping cues, reversed timings, and unparseable blocks, is reported as a warning instead of stopping the conversion
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/discount-calculator`, `POST /v1/tools/subtitle-converter`), Node SDK (`calculateDiscount` / `applySuccessiveDiscounts`, `convertSubtitles` / `shiftSubtitles` / `parseSubtitles`), Python SDK (`calculate_discount` / `apply_successive_discounts`, `convert_subtitles` / `shift_subtitles` / `parse_subtitles`), and MCP server (`calculate_discount`, `convert_subtitles`)
- Node SDK bumped to v0.24.0, Python SDK to v0.24.0, MCP server to v0.25.0 (137 tools total, up from 135)
July 25, 2026New Tools
v0.23.0 / v0.23.0 / v0.24.0Margin & Markup Calculator + SQL Query Explainer (AI)
- Margin & Markup Calculator: fill in any two of cost, selling price, gross margin percent, and markup percent, and the other two are solved for you along with the gross profit per unit. Margin is profit / price and markup is profit / cost, so an item costing 80 sold at 100 is a 20% margin and a 25% markup from the same 20 of profit: conflating the two is the mistake this page exists to prevent
- SQL Query Explainer (AI): paste a SQL query and get a plain-English summary of what it returns, an ordered walkthrough of how it is evaluated, every table and CTE with the role it plays, and performance patterns worth checking, powered by GPT-4o mini. Pin a dialect (Postgres, MySQL, SQLite, SQL Server, Oracle, BigQuery, Snowflake) or let the model infer it from the syntax. It reads the query text only, with no access to your schema, indexes, or a real query plan, so the performance notes are leads to confirm with EXPLAIN
- Margin & Markup Calculator shipped across every surface: browser UI, REST API (`POST /v1/tools/margin-markup-calculator`), Node SDK (`calculateMarginMarkup`, plus `marginToMarkup` / `markupToMargin` helpers), Python SDK (`calculate_margin_markup`, `margin_to_markup` / `markup_to_margin`), and MCP server (`calculate_margin_markup`)
- SQL Query Explainer is browser + REST only (`POST /v1/ai/sql-explain`), like the other LLM-backed tools. Free plan includes 3 explanations per month, Pro includes 25, and usage shows on your dashboard
- Node SDK bumped to v0.23.0, Python SDK to v0.23.0, MCP server to v0.24.0 (135 tools total, up from 134)
July 22, 2026New Tools
v0.22.0 / v0.22.0 / v0.23.0APR / APY Converter + ISO 8601 Duration Parser
- APR / APY Converter: convert between a nominal annual rate (APR) and the effective annual yield (APY) for any compounding frequency, in both directions. APR to APY uses (1 + APR/n)^n - 1, and the reverse recovers the APR, so a 12% APR compounded monthly reads as a 12.682503% APY
- ISO 8601 Duration Parser / Formatter: parse a duration string like P3DT4H30M into its components and a total number of seconds, or format a number of seconds back into a canonical duration. Years and months use nominal 365-day / 30-day conversions and flag the total as approximate, since they have no fixed length
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/apr-apy-converter`, `POST /v1/tools/iso8601-duration`), Node SDK (`convertAprApy`, `parseIso8601Duration` / `formatIso8601Duration`), Python SDK (`convert_apr_apy`, `parse_iso8601_duration` / `format_iso8601_duration`), and MCP server (`convert_apr_apy`, `convert_iso8601_duration`)
- Node SDK bumped to v0.22.0, Python SDK to v0.22.0, MCP server to v0.23.0 (134 tools total, up from 132)
July 21, 2026New Tools
v0.21.0 / v0.21.0 / v0.22.0Percentage Change Calculator + JSON Pointer Resolver
- Percentage Change Calculator: compute the percentage increase or decrease between two numbers as ((new - old) / |old|) x 100, plus the absolute change, the direction (increase/decrease/no change), and the multiplier (new / old). Divides by the original value the way the standard formula requires, so 200 to 250 reads as +25% rather than the +20% you get dividing by the new value
- JSON Pointer Resolver: resolve an RFC 6901 JSON Pointer (like /users/0/name) against a document, handling the ~1 (slash) and ~0 (tilde) escapes, the empty-pointer whole-document case, and the #/... URI-fragment form, and reporting exactly which segment failed when a target is absent
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/percentage-change-calculator`, `POST /v1/tools/json-pointer-resolver`), Node SDK (`calculatePercentageChange`, `resolveJsonPointer`), Python SDK (`calculate_percentage_change`, `resolve_json_pointer`), and MCP server (`calculate_percentage_change`, `resolve_json_pointer`)
- Node SDK bumped to v0.21.0, Python SDK to v0.21.0, MCP server to v0.22.0 (132 tools total, up from 130)
July 20, 2026New Tools
v0.20.0 / v0.20.0 / v0.21.0Cache-Control Header Builder + Filler Word & Cliché Detector
- Cache-Control Header Builder: assemble an HTTP `Cache-Control` response header from directives (public/private, no-cache, no-store, max-age, s-maxage, must-revalidate, immutable, stale-while-revalidate, and more), with a plain-English explanation of each directive and warnings for combinations that cancel out, like pairing no-store with a max-age
- Filler Word & Cliché Detector: scan prose for filler words (just, really, basically), hedge phrases (kind of, I think), and clichés (at the end of the day, circle back), returning per-phrase counts, character offsets, and an overall density per 100 words
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/cache-control-builder`, `POST /v1/tools/filler-word-detector`), Node SDK (`buildCacheControl`, `detectFillerWords`), Python SDK (`build_cache_control`, `detect_filler_words`), and MCP server (`build_cache_control`, `detect_filler_words`)
- Node SDK bumped to v0.20.0, Python SDK to v0.20.0, MCP server to v0.21.0 (130 tools total, up from 128)
July 19, 2026New Tools
v0.19.0 / v0.19.0 / v0.20.0Regex ReDoS Detector + Dockerfile Generator (AI)
- Regex ReDoS Detector: statically scan a regex for catastrophic-backtracking risk, flagging nested quantifiers like (a+)+ (high, exponential), overlapping alternation under a repeat like (a|ab)* (medium), and adjacent overlapping quantifiers like \d+\d+ (medium, quadratic), while leaving safe patterns like (ab)+ and (a|b)+ alone
- Dockerfile Generator (AI): describe an app in plain English and get a production-ready Dockerfile back (pinned base image, multi-stage builds, layer caching, non-root user), powered by GPT-4o mini. Free: 3/month, Pro: 25/month
- Regex ReDoS Detector shipped across every surface: browser UI, REST API (`POST /v1/tools/redos-detector`), Node SDK (`detectRedos`), Python SDK (`detect_redos`), and MCP server (`detect_redos`)
- Dockerfile Generator ships as a browser tool + REST API (`POST /v1/ai/dockerfile`)
- Node SDK bumped to v0.19.0, Python SDK to v0.19.0, MCP server to v0.20.0 (128 tools total, up from 127)
July 18, 2026New Tools
v0.18.0 / v0.18.0 / v0.19.0Present Value / Future Value Calculator + Vector Similarity Calculator
- Present Value / Future Value Calculator: convert between present and future value at a given rate in either direction (grow a present sum to its future value, or discount a future sum back to today), with annual/semi-annual/quarterly/monthly/daily compounding
- Vector Similarity Calculator: compute cosine similarity, dot product, and Euclidean / Manhattan distance between two embedding vectors, for debugging RAG pipelines that already have vectors from an external model
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/present-value-calculator`, `POST /v1/tools/vector-similarity`), Node SDK (`calculateTimeValue`, `vectorSimilarity`), Python SDK (`calculate_time_value`, `vector_similarity`), and MCP server (`calculate_time_value`, `vector_similarity`)
- Node SDK bumped to v0.18.0, Python SDK to v0.18.0, MCP server to v0.19.0 (127 tools total, up from 125)
July 15, 2026New Tools
v0.17.0 / v0.17.0 / v0.18.0User-Agent Parser + GitHub URL Parser
- User-Agent Parser: parse a browser User-Agent string into browser name/version, rendering engine, OS, and device type (desktop, mobile, tablet, or bot) via pattern matching
- GitHub URL Parser: parse a github.com URL (repo, pull request, issue, commit, file, directory, or release tag) into structured components, including a #L10-L20 line-range anchor for file URLs
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/user-agent-parser`, `POST /v1/tools/github-url-parser`), Node SDK (`parseUserAgent`, `parseGithubUrl`), Python SDK (`parse_user_agent`, `parse_github_url`), and MCP server (`parse_user_agent`, `parse_github_url`)
- Node SDK bumped to v0.17.0, Python SDK to v0.17.0, MCP server to v0.18.0 (125 tools total, up from 123)
July 14, 2026New Tools
v0.16.0 / v0.16.0 / v0.17.0Simple Interest Calculator + Base58 / Base62 Encoder
- Simple Interest Calculator: compute simple (non-compounding) interest and total payoff from principal, annual rate, and term, as a distinct sibling to the Compound Interest Calculator
- Base58 / Base62 Encoder: encode and decode text as Base58 (Bitcoin-address style, no ambiguous 0/O/I/l characters) or Base62 (alphanumeric, URL-safe)
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/simple-interest-calculator`, `POST /v1/tools/base58-base62-encoder`), Node SDK (`calculateSimpleInterest`, `base58Encode`/`base58Decode`/`base62Encode`/`base62Decode`), Python SDK (`calculate_simple_interest`, `base58_encode`/`base58_decode`/`base62_encode`/`base62_decode`), and MCP server (`calculate_simple_interest`, `base58_encode`/`base58_decode`/`base62_encode`/`base62_decode`)
- Node SDK bumped to v0.16.0, Python SDK to v0.16.0, MCP server to v0.17.0 (123 tools total, up from 118)
July 13, 2026New Tools
v0.15.0 / v0.15.0 / v0.16.0MIME Type Lookup + Tip / Bill Split Calculator
- MIME Type Lookup: look up the MIME type for a file extension, or the file extension(s) registered for a MIME type, from a static table of well-known, IANA-registered content types
- Tip / Bill Split Calculator: compute a tip on a bill amount and split the bill, tip, and total evenly across any number of people
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/mime-lookup`, `POST /v1/tools/tip-calculator`), Node SDK (`lookupMime`, `calculateTipSplit`), Python SDK (`lookup_mime`, `calculate_tip_split`), and MCP server (`lookup_mime`, `calculate_tip_split`)
- Node SDK bumped to v0.15.0, Python SDK to v0.15.0, MCP server to v0.16.0 (118 tools total, up from 116)
July 12, 2026New Tools
v0.14.0 / v0.14.0 / v0.15.0Compound Interest Calculator + ISBN Validator / Converter
- Compound Interest Calculator: compute the future value of a principal plus an optional recurring contribution under compound interest, given an annual rate, compounding frequency (annually/quarterly/monthly/daily), and term in years, with a year-by-year balance breakdown
- ISBN Validator / Converter: validate an ISBN-10 (mod-11 checksum, including the X check digit) or ISBN-13 (weighted mod-10 checksum) and convert a valid ISBN to the other format where possible
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/compound-interest-calculator`, `POST /v1/tools/isbn-validator`), Node SDK (`calculateCompoundInterest`, `validateIsbn`), Python SDK (`calculate_compound_interest`, `validate_isbn`), and MCP server (`calculate_compound_interest`, `validate_isbn`)
- Node SDK bumped to v0.14.0, Python SDK to v0.14.0, MCP server to v0.15.0 (116 tools total, up from 114)
July 11, 2026Fix
AI tools let you re-enter a key when yours stops working
- Every AI-powered tool (SVG Generator, NL → Regex, CLI Builder, CSS Explainer, JSON Schema Generator, Commit Message Generator, Stack Trace Summarizer) now reopens the API-key input when a stored key is revoked or expired, instead of showing an error with no way to recover
- Paste a fresh key from your dashboard and keep going, no reload or clearing site data needed
July 11, 2026New Tools
v0.13.0 / v0.13.0 / v0.14.0VAT / Sales Tax Calculator + Stack Trace Summarizer (AI)
- VAT / Sales Tax Calculator: add tax to a net price or remove tax from a gross price in either direction, with a reference table of common VAT/GST/sales-tax rates for convenience, entirely client-side
- Stack Trace Summarizer (AI): paste an error stack trace and get the likely error type, a plain-English explanation of the cause, the most relevant frame, and a suggested fix, powered by GPT-4o mini
- VAT / Sales Tax Calculator shipped across every surface: browser UI, REST API (`POST /v1/tools/vat-calculator`), Node SDK (`calculateTax`), Python SDK (`calculate_tax`), and MCP server (`calculate_tax`)
- Stack Trace Summarizer shipped to browser UI and REST API (`POST /v1/ai/stack-trace`), matching the existing AI tools' free/Pro usage limits (3/month free, 25/month Pro)
- Node SDK bumped to v0.13.0, Python SDK to v0.13.0, MCP server to v0.14.0 (114 tools total, up from 113)
July 10, 2026New Tools
v0.12.0 / v0.12.0 / v0.13.0ICO Favicon Inspector + Barcode Checksum Validator
- ICO Favicon Inspector: drop in a .ico file and see how many images it embeds and each one's width, height, color depth, and byte size, read directly from the ICONDIR/ICONDIRENTRY header table with no image decoding, and rejecting .cur cursor files with a clear error
- Barcode Checksum Validator: validate an EAN-13, EAN-8, or UPC-A barcode via its weighted mod-10 check digit algorithm, with the format inferred automatically from the digit count
- Shipped across every surface: browser UI, REST API (`POST /v1/tools/ico-favicon-inspector`, `POST /v1/tools/barcode-validator`), Node SDK (`readIcoInfo`, `validateBarcode`), Python SDK (`read_ico_info`, `validate_barcode`), and MCP server (`read_ico_info`, `validate_barcode`)
- Node SDK bumped to v0.12.0, Python SDK to v0.12.0, MCP server to v0.13.0 (113 tools total, up from 111)
July 9, 2026New Tools
v0.11.0 / v0.11.0 / v0.12.0IBAN Validator / Formatter + Loan / Mortgage Calculator
- IBAN Validator / Formatter: validate an IBAN via the mod-97 checksum defined by ISO 13616, cross-check its length against the registered length for ~80 countries, and format it into the standard grouped form, entirely client-side
- Loan / Mortgage Calculator: compute the fixed monthly payment, total interest, and a full month-by-month amortization schedule from principal, annual rate, and term, using the standard amortizing-loan formula (0% rate handled as a straight-line split)
- Shipped across every surface: browser UI, REST API (`POST /v1/tools/iban-validator`, `POST /v1/tools/loan-calculator`), Node SDK (`validateIban`, `calculateLoan`), Python SDK (`validate_iban`, `calculate_loan`), and MCP server (`validate_iban`, `calculate_loan`)
- Node SDK bumped to v0.11.0, Python SDK to v0.11.0, MCP server to v0.12.0 (111 tools total, up from 109)
July 8, 2026New Tools
v0.10.0 / v0.10.0 / v0.11.0robots.txt Validator + Credit Card Number Validator
- robots.txt Validator: paste a robots.txt file and see it parsed into User-agent groups, with structural mistakes flagged: directives before any User-agent, unknown directives, non-absolute Sitemap URLs, malformed Crawl-delay values, missing leading slashes, and duplicate rules
- Credit Card Number Validator: validate a card number via the Luhn checksum and detect its network (Visa, Mastercard, American Express, Discover, Diners Club, JCB, Maestro, UnionPay) from prefix and length, entirely client-side with no card data ever stored or transmitted
- Shipped across every surface: browser UI, REST API (`POST /v1/tools/robots-txt-validator`, `POST /v1/tools/credit-card-validator`), Node SDK (`validateRobotsTxt`, `validateCardNumber`), Python SDK (`validate_robots_txt`, `validate_card_number`), and MCP server (`validate_robots_txt`, `validate_card_number`)
- Node SDK bumped to v0.10.0, Python SDK to v0.10.0, MCP server to v0.11.0 (109 tools total, up from 107)
July 7, 2026New Tools
v0.9.0 / v0.9.0 / v0.10.0Sitemap.xml Validator
- Paste a sitemap.xml (urlset) or sitemap index (sitemapindex) and get every structural issue flagged with a line number: missing or relative `<loc>`, unescaped ampersands, out-of-range `<priority>`, an invalid `<changefreq>` value, duplicate URLs, and the 50,000-URL cap
- Softer issues (missing XML declaration, non-standard namespace, unparseable `<lastmod>`) are reported as warnings rather than errors, so it's clear which problems block crawling and which are just untidy
- Shipped across every surface: browser UI, REST API (`POST /v1/tools/sitemap-validator`), Node SDK (`validateSitemap`), Python SDK (`validate_sitemap`), and MCP server (`validate_sitemap`)
- Node SDK bumped to v0.9.0, Python SDK to v0.9.0, MCP server to v0.10.0 (107 tools total, up from 106)
July 7, 2026Breaking Change
File-upload REST endpoints switched to multipart/form-data
- `POST /v1/tools/image-info`, `pdf-metadata`, `exif-viewer`, `wav-audio-info`, and `id3-tag-reader` now take the file directly as `multipart/form-data` (a single `file` field) instead of a base64-encoded string in a JSON body
- This is a breaking change with no transition period: `{"mp3Base64": "..."}`-style JSON bodies are no longer accepted on these 5 endpoints — send the raw file with `curl -F "file=@track.mp3"` or your HTTP client's multipart/form-data support instead
- Fixes every affected tool page's REST API example and the API Playground, both of which previously showed a broken example for these endpoints
July 7, 2026New Tools
v0.8.0 / v0.8.0 / v0.9.0ID3 Tag Reader + .env.example Diff Checker
- ID3 Tag Reader: read title, artist, album, year, genre, comment, and track number directly from an MP3 file's bytes — prefers an ID3v2.3/2.4 header, falls back to the classic ID3v1/1.1 128-byte trailer, no audio library required
- .env.example Diff Checker: paste a `.env` file and its `.env.example` template to see keys missing from either side, plus keys left with an empty placeholder value in the example — catches the classic "new teammate's .env is silently missing a key" onboarding failure
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/id3-tag-reader`, `POST /v1/tools/env-diff-checker`), Node SDK, Python SDK, and MCP server (`read_id3_tags`, `diff_env_example`)
- Node SDK bumped to v0.8.0, Python SDK to v0.8.0, MCP server to v0.9.0 (106 tools total, up from 104)
July 6, 2026New Tools
v0.7.0 / v0.7.0 / v0.8.0WAV Audio Info + Readability Score Calculator
- WAV Audio Info: read sample rate, channel count, bit depth, byte rate, and duration directly from a WAV file's RIFF/fmt/data chunk headers — pure byte parsing, no audio library, canonical RIFF/WAVE container only
- Readability Score Calculator: paste text and get Flesch Reading Ease, Flesch-Kincaid Grade Level, and Gunning Fog Index back at once, using a heuristic syllable counter — no dictionary lookup required
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/wav-audio-info`, `POST /v1/tools/readability-score`), Node SDK, Python SDK, and MCP server (`read_wav_info`, `score_readability`)
- Node SDK bumped to v0.7.0, Python SDK to v0.7.0, MCP server to v0.8.0 (104 tools total, up from 102)
July 5, 2026New Tools
v0.6.0 / v0.6.0 / v0.7.0EXIF Viewer + Commit Message Generator (AI)
- EXIF Viewer: read camera make/model, orientation, timestamps, exposure settings, and GPS coordinates directly from a JPEG's APP1 segment — pure byte parsing, no image library, JPEG only
- Commit Message Generator (AI): paste a git diff and get a Conventional Commits-style message back (type, scope, subject, body), powered by GPT-4o mini — Free: 3/month, Pro: 25/month
- EXIF Viewer shipped across every surface: browser UI, REST API (`POST /v1/tools/exif-viewer`), Node SDK, Python SDK, and MCP server (`read_exif_data`)
- Commit Message Generator shipped as browser UI + REST API (`POST /v1/ai/commit-message`), matching the existing 5 AI-generation tools' surface pattern (no SDK/MCP wrapper, since it requires a Pro-tier API key and per-user rate limiting)
- Node SDK bumped to v0.6.0, Python SDK to v0.6.0, MCP server to v0.7.0 (102 tools total, up from 101)
July 4, 2026New Tools
v0.5.0 / v0.5.0 / v0.6.0PDF Metadata Reader + JWK Viewer
- PDF Metadata Reader: read title, author, subject, keywords, creator, producer, creation/mod dates, page count, and encryption flag directly from a PDF's trailer/Info dictionary — no pdf.js or other PDF library, pure byte parsing
- JWK Viewer: parse a JSON Web Key Set (or a single bare JWK) and view each key's type, use, algorithm, key ID, and approximate size in a readable table, with a duplicate-kid warning for debugging JWT verification failures
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/pdf-metadata`, `POST /v1/tools/jwk-viewer`), Node SDK, Python SDK, and MCP server (`read_pdf_metadata`, `parse_jwks`)
- Node SDK bumped to v0.5.0, Python SDK to v0.5.0, MCP server to v0.6.0 (101 tools total, up from 99)
July 3, 2026Fix
Five remaining consistency bugs from the site audit
- Pricing FAQ still said "100+ browser tools" while the feature list/table already said the real count — wired the FAQ to the same shared constant so it can never drift again
- Docs page showed "400+ functions" in the Quick start card but "510+/520+" in the detailed Node.js SDK section, on the same page — both now read from one source
- MCP category breakdown table summed to 70 while the headline said 97 (now 99) — recomputed real per-category counts directly from the MCP tool registry (99 tools: JSON 13, Encoding 8, Hashing 4, Generators 5, Text 11, Time 5, Units 2, Network/API 5, Color 4, Code 7, Data 5, CSS 2, Misc 1, AI Agent 27)
- /api Playground upsell banner had a math error (13 + 82 ≠ 100) and an undercounted total: the REST API actually exposes 104 tool endpoints (apps/api/src/routes/tools.ts + tools2.ts), not 100. Also fixed 3 duplicate entries in the Playground's own endpoint list (repair-json, sanitize-html, extract-urls were each listed twice)
- Added a regression test asserting the upsell banner (and its "Full docs" CTA) is hidden when signed in and shown when signed out
July 3, 2026New Tools
v0.4.0 / v0.4.0 / v0.5.0HAR File Viewer + NDJSON Formatter / Validator
- HAR File Viewer: parse a browser DevTools .har network export into a readable table of requests, status codes, sizes, and timing, plus aggregate summary stats (total size, total/avg time, failed request count)
- NDJSON Formatter / Validator: validate newline-delimited JSON (NDJSON/JSON Lines) with per-line error reporting, pretty-print or minify each line, and convert between NDJSON and a JSON array
- Both tools shipped across every surface: browser UI, REST API (`POST /v1/tools/har-parse`, `POST /v1/tools/ndjson-format`), Node SDK, Python SDK, and MCP server (`parse_har`, `format_ndjson`)
- Node SDK bumped to v0.4.0, Python SDK to v0.4.0, MCP server to v0.5.0 (99 tools total, up from 97)
July 2, 2026SDK Fix
v0.3.1 / v0.3.1 / v0.4.1Node SDK v0.3.1, Python SDK v0.3.1, MCP v0.4.1: npm/PyPI listing fixes
- Fixed a broken link on the PyPI package page: Repository and Bug Tracker pointed to github.com/utilix-tech/utilix-sdk-python, a repo that doesn't exist. Corrected to the real utilix-sdk repo
- MCP server was pinned to @utilix-tech/sdk ^0.2.0 while the SDK had already moved to 0.3.0 — bumped the floor to ^0.3.0 to close a gap where a future tool addition could silently resolve to a stale SDK version
- Added missing LICENSE file to the MCP package (was declared MIT but never actually published)
- Added CHANGELOG.md and SECURITY.md to all three packages (Node, Python, MCP)
- Added repository/bugs fields to Node and MCP package.json (previously missing from the npm listing entirely)
- Expanded keywords for discoverability, and corrected stale "100+" tool-count claims in package descriptions to 140+
July 2, 2026Fix
Sitewide consistency, SEO, and accessibility audit
- Fixed a stale "100 req/day" rate-limit claim on every tool page's code snippet (real limit is 1,000/day) — the exact bug this audit was meant to catch
- Reconciled MCP tool count (97), Node SDK function count (510+), Python SDK function count (380+), and total browser tool count (141+) everywhere they appeared, and centralized them in a single apps/web/src/lib/stats.ts so they can't drift apart again
- og:description and twitter:description were identical generic text on every page — now every page (pricing, docs, tools, individual tool pages, changelog, legal pages, API Playground) sets its own
- Added "How to use" + FAQ content (with FAQPage structured data) to the 10 highest-traffic tool pages: JSON Formatter, JWT Decoder, Hash Generator, Base64, URL Encode, UUID Generator, Regex Tester, Color Picker, Cron Parser, and CSS Gradient
- Pricing table's checkmark/dash cells now have screen-reader text ("Included"/"Not included") instead of being icon-only
- Team plan CTA changed from "Contact us" to "Join waitlist" so it doesn't imply the plan is sales-ready yet
July 2, 2026Fix
API Playground demo mode was missing 3 live endpoints
- word-counter, unix-time, and px-rem were live on the backend but never listed in the demo tool picker: fixed to show all 13
- Pricing page updated from "10 endpoints" to "13 endpoints" to match
July 2, 2026Site
"Recently added" tools rail and New filter
- Homepage now shows the 5 most recently added tools in a dedicated "Recently added" section
- /tools gets a New filter: toggle it to see just the newest tools, separate from category browsing, without losing access to category browsing (dimmed, not hidden, while active)
- The rail rotates automatically as new tools ship: oldest entries drop off once 5 are flagged
- Homepage "Browse by category" now surfaces the 6 categories with the most tools (AI/Agent, Network, JSON, CSS, Code, Text) instead of a fixed list, so it stays accurate as the catalog grows
- Fixed /tools?category=x links from the homepage to actually scroll to and highlight that category
July 2, 2026New Tools
v0.4.0Passive Voice Detector + Image Info, plus a critical MCP fix
- New tool: Passive Voice Detector: flags sentences written in passive voice and highlights the exact phrase that triggered each match
- New tool: Image Info: reads format, dimensions, bit depth, and alpha channel straight from image file bytes (PNG, JPEG, GIF, WebP, BMP), no decoding required
- Both tools shipped across the browser, REST API, Node.js SDK, Python SDK, and MCP server: 97 MCP tools total
- Node.js SDK bumped to v0.3.0 (new media module), Python SDK to v0.3.0, MCP server to v0.4.0
- Fixed a critical MCP bug: the published @utilix-tech/mcp package depended on an SDK version older than the one that introduced the ai_agent module, so the server was crashing at startup for all 25 ai_agent-derived tools. The dependency was bumped and the fix was verified
July 2, 2026MCP Fix
v0.3.1Fixed silently-broken diff_json MCP tool
- MCP server had two tools both registered as diff_json: the second registration silently overrode the first, making the richer path/counts diff unreachable
- Renamed the object-based variant to diff_json_detailed (adds unchanged-path reporting, useful for agent diffing workflows): both diff tools are now reachable
- MCP tool count corrected to 95 (was previously miscounted as 94 due to the collision)
July 2, 2026Blog + Tools
Blog launch, API Playground, /mcp page, and two new tools
- Blog launched at /blog: deep-dive guides on AI agents, MCP, and developer tooling, with search, tag/category filtering, and related-post recommendations
- API Playground at /api: try 10 REST endpoints live with no signup, full access to 100+ endpoints with an API key
- Dedicated /mcp page: one-command install instructions per editor, all 94 MCP tools listed by category, example prompts
- New browser tools: SSL Certificate Checker and Email Validator
- MCP server verified at 94 tools (previously undercounted as 70 in docs because the ai_agent module additions from v0.3.0 were missing from public-facing tool lists)
June 30, 2026MCP + SDK
v0.3.0SDK v0.2.0 · MCP v0.3.0 · Python SDK
- Node.js SDK bumped to v0.2.0: 400+ functions across 14 modules
- Python SDK v0.2.0: 351 functions across 16 modules, identical return shapes to Node
- MCP server expanded to 95+ tools (up from 70)
- Added full ai_agent module: 28 tools for LLM pipelines (token management, compression, PII, RAG, JSON utilities)
- SDK examples reorganized into examples/node/ and examples/python/ with verified working code
- Added 60+ example prompts for MCP (prompts.md on GitHub)
- Docs: new Examples section linking all public repos
June 29–30, 2026AI Tools
15 new AI agent tools
- Token Estimator: count tokens and estimate LLM cost before sending
- PII Detector: find emails, phones, SSNs, credit cards in text
- Secret Detector: scan for leaked API keys, tokens, passwords
- Prompt Injection Detector: score and flag adversarial inputs
- JSON Repair: fix malformed JSON with a list of applied fixes
- Context compression: HTML, Markdown, JSON compressors
- RAG utilities: chunk text, rerank chunks, expand query, summarize
- Extraction tools: extract URLs, JSON blocks, entities from text
June 28, 2026Launch
v0.1.0Public launch
- 107 browser tools across 14 categories: no login required, runs client-side
- REST API at api.utilix.tech/v1 with interactive reference at api.utilix.tech/docs
- Node.js SDK v0.1.0 (@utilix-tech/sdk) on npm
- MCP Server v0.2.0 (@utilix-tech/mcp): works with Claude Desktop, Cursor, VS Code Copilot
- Pro plan ($9/month) with Paddle: 10k API req/day, AI tools, snippet manager
- AI tools: SVG Generator, NL→Regex, CLI Builder, CSS Explainer, JSON Schema Generator
- Dashboard with API key management and usage charts
- Snippet manager: save and retrieve tool outputs via API