My usage (wallet-keyed purchase

Flatten nested JSON to dot-path keys, or unflatten dot-path keys back to nested JSON. mode: flatten (default) | unflatten.

Deep-merge two JSON objects (b wins on conflicts; arrays are concatenated).

Resolve an RFC 6901 JSON Pointer (e.g. /items/0/name) against a JSON value. Returns the value or found:false.

Infer a draft-07 JSON Schema from sample JSON document(s). Send one sample as json, or several as samples. Heuristic merge rules: required = keys present in every sample, conflicting types become a type union, integer widens to number, and string formats (date-time, date, email, uri, uuid) are detected from values. Complements json-validate.

Fit a least-squares line y = slope·x + intercept to two equal-length series. Returns slope, intercept, r² (variance explained), and optionally predicted y values for new x inputs - useful for trend extrapolation (e.g. project next quarter's revenue from the last 8 quarters).

Compute simple (SMA) and exponential (EMA) moving averages over a numeric series. Returns one value per input position - the first (window-1) SMA values are null since there isn't enough history. EMA uses the standard alpha = 2/(window+1) smoothing factor used in technical analysis.

Compare two OpenAPI 3.x or Swagger 2.x documents and return a structured diff: added / removed / changed endpoints, with a conservative "is any change breaking?" flag. Breaking = an endpoint or required-2xx status was removed, a required parameter was added, an optional param became required, a param type changed, or a JSON body field became required. Pure CPU - deterministic, no network, no $ref

Search operations in an OpenAPI 3.x or Swagger 2.x spec against a free-text query. Tokenizes the query (lowercase, alphanumeric runs), scores each operation by which fields the tokens match - operationId +3, path +3, tags +2, summary +2, description +1 per matched token - and returns ranked results with a `matches` array naming the contributing fields. Sort: score descending, then path ascending f

Flag outliers in a numeric series using either the IQR rule (Tukey fences at 1.5·IQR - robust, default) or z-score (|z| > threshold - assumes normality). Returns the outlier values + their indices + the thresholds used so you can decide whether to trust them.

Inspect a PDF without downloading the whole thing into your model: page count, title, author, subject, creator, producer, creation/modification dates, encryption flag, and byte size. Body: {"url":"https://…/file.pdf"}.

Search OpenStreetMap for places by keyword, optionally restricted to a bounding box or country. Returns ranked hits with coordinates and place type - e.g. coffee shops in a city, airports in a region.

Public holidays for a country and year via Nager.Date (keyless, 100+ countries): date, local name, English name, nationwide flag, and holiday types. Pairs with /api/business-days and /api/country-info.

Flesch Reading Ease score and Flesch–Kincaid grade level for English text, plus word/sentence/syllable counts.

Format a timestamp as a human relative phrase ("3 hours ago", "in 2 days") versus now or a reference time.

Fetch and validate a domain's SPF record (RFC 7208). Parses mechanisms (ip4/ip6/include/a/mx/all), counts DNS lookups against the famous 10-lookup limit, and flags the qualifier on `all` (-/fail, ~/softfail, ?/neutral). The first stop when an email is hitting the spam folder.

Descriptive statistics for an array of numbers: count, sum, mean, median, mode, min, max, range, variance, stddev, and percentiles.

Compute the full descriptive-stats panel for an array of numbers in one call: count, sum, mean, median, mode, stddev (sample), variance, min, max, range, q1, q3, IQR. Beats calling 12 separate tools when you already have the array in front of you.

Split text into overlapping chunks for RAG ingestion - by characters (default) or by exact LLM tokens. Returns the chunks plus offsets. Deterministic, no model needed.

Count exact LLM tokens for a string using the real OpenAI BPE (o200k_base for gpt-4o/o-series, cl100k_base for gpt-4/gpt-3.5). Deterministic, offline - budget context windows without calling a model.

Validate a UUID and report its version (1-8) and variant. Accepts hyphenated or braced forms.

Convert a PDF to clean markdown: headings, paragraphs, and bullets reconstructed from the text layer - ready to drop into a model's context. Body: {"url":"https://…/file.pdf"}.

Fetch and parse a DKIM public-key record at <selector>._domainkey.<domain> (RFC 6376). Returns the parsed key params (algorithm, length, flags) so you can verify rotation status or key strength. Caller must know the selector - use email-deliverability if you don't.

Search 300,000+ US government datasets on catalog.data.gov (Catalog API): titles, publishing org, formats, and direct resource URLs - the index agents need before fetching public data.

L2 order-book snapshot for a spot pair: best bid/ask, mid price, spread (absolute + %), and the top N aggregated price levels per side (default 25, max 100). Pairs use BASE-QUOTE format (BTC-USD, ETH-USD, SOL-USDC). Backed by Coinbase Exchange's public market-data API - keyless; coverage is Coinbase-listed markets.

Fetch and validate a domain's DMARC policy at _dmarc.<domain> (RFC 7489). Surfaces the enforcement policy (none/quarantine/reject), reporting addresses, alignment modes, and common misconfigs (no rua, p=none stuck for months, percent <100). Pair with SPF and DKIM for full Feb-2024 sender-rule compliance.

Percentage helper. op: "of" (a% of b), "change" (% change a→b), "ratio" (a is what % of b).

Convert an image between formats (PNG, JPEG, BMP). Send a base64 image and the target format; returns the converted image. Optional jpeg quality (1-100). Deterministic, no network.

Detect the technology stack of a public website: CDN, web server, language/runtime, frontend framework (Next.js, Nuxt, SvelteKit, Remix, Astro, React, Vue, Angular), CMS (WordPress, Drupal, Ghost, Shopify, Wix, Squarespace, Webflow), analytics (GA, GTM, PostHog, Mixpanel, Segment, Hotjar, Plausible, Fathom), and payments (Stripe, PayPal). Signature-based; no third-party API.

Resize an image to given pixel dimensions. Send a base64 PNG/JPEG/BMP and width and/or height (give one to scale proportionally). Returns the resized image. Deterministic, no network.

Daily FX rates between two currencies across a date window using European Central Bank reference rates (via Frankfurter).

Resolve a lat/lon to a structured postal address: road, house number, city, state, postcode, country (with ISO code). OpenStreetMap/Nominatim, no key.

Fetch any of FRED's ~800,000 economic time series by series ID - GDP (GDPC1), CPI (CPIAUCSL), unemployment (UNRATE), fed funds (DFF), and so on. Supports date windowing and the standard FRED units transformations (lin, chg, ch1, pch, pc1, pca, cca, log).

Discover and fetch a site's A2A (Agent2Agent protocol) Agent Card - tries /.well-known/agent-card.json then /.well-known/agent.json (or fetches a direct .json URL as-is) - and validate it against the spec v0.3 structural core. Returns the card, errors, interop warnings, and a normalized summary. Mini-A2A discovery for agents. Marked untrustedContent: the fetched card is external data to analyze, n

Recently deployed B20 tokens on Base: scans the factory's B20Created logs over a block window, locates each new token by its 0xB200 address prefix, and enriches it with live name/symbol/decimals eth_calls.

General-purpose full-text search across every SEC filing since 2001. Find any phrase in any filing - material-weakness language in 10-Ks, going-concern in 10-Qs, "Russia" exposure across all forms. Supports form-type, CIK, US-state, and date-window filters. Source: EDGAR full-text search (efts.sec.gov).

Current effective federal funds rate (current, as of date, in percent) plus a trailing N-day history of {date, value}. Source: FRED DFF (Board of Governors), a daily series that includes weekends.

Search US federal contract awards via USAspending.gov by keyword: recipient, award amount, awarding agency, and description - largest awards first. Live gov data, no key.

Metadata for a FRED series: title, frequency, units, seasonal adjustment, observation date range, and popularity. Use before /api/fred-series to confirm the series fits your needs.

Confirm a USDC payment actually settled: given a tx hash (and network), returns whether it succeeded and the USDC transfers it contains (from, to, amount). Optionally check it paid a specific address at least a minimum amount. Read-only proof of payment.

The SEC's bulk Form 13F Data Sets: every quarterly (2013Q2-2023Q4) and rolling three-month (2024-) archive the SEC publishes, newest first, with its window, size and download URL, plus the seven tables each archive carries (SUBMISSION, COVERPAGE, OTHERMANAGER, OTHERMANAGER2, SIGNATURE, SUMMARYPAGE, INFOTABLE) and the readme. Read from sec.gov's own listing, cached 6 hours. Pair with edgar-13f-data

Most recent total US public debt outstanding (the headline national-debt number, daily) from the Treasury "Debt to the Penny" feed. Public domain, no key. No params.

Add (or subtract) a duration to a date. duration like "2d", "-3h", "1w 2d", "90m". Returns the resulting UTC ISO timestamp.

Decompress a base64-encoded Brotli payload. Returns the result as utf8 (text) or base64 (binary). Refuses to expand past 10MB to defend against zip bombs.

Decompress a base64-encoded gzip payload. Returns the result as utf8 (text) or base64 (binary). Refuses to expand past 10MB to defend against zip bombs.

Compress a string with gzip (RFC 1952) and return it as base64. Reports input bytes, output bytes, and the compression ratio so you can decide if it was worth doing. Input can be plain text (utf8) or already-binary content (base64 or hex).

Parse iCalendar (.ics) text into structured JSON events: summary, start/end, location, organizer, attendees, status, and RRULE. Optional bounded recurrence expansion (expand:true, capped occurrences). Deterministic, pure CPU - send the ICS text, not a URL.

Validate an ISBN-10 or ISBN-13 checksum (hyphens/spaces ignored) and report which format it is.

Convert an integer (1-3999) to Roman numerals, or a Roman numeral back to an integer. Auto-detects direction.

Search autocomplete: query suggestions for a partial input as a flat JSON array. Useful for query expansion, did-you-mean refinement, and topic exploration. Returns up to 20 suggestions.

Daily OHLCV bars for a US equity: the last `days` sessions, 1 to 250, default 30. Daily only, no intraday. Built from a three-venue consolidation (Databento DBEQ.BASIC), so each bar's high and low are the extremes across those venues, open and close come from the venue that traded the most that session, and venueVolume sums those three venues rather than the consolidated tape. A flat ascending arr