# divux (remote · mcp.divux.si) Official data for AI assistants: medicines, recalls, cosmetics, money, travel, places, prices. - Trust score: 73/100 (medium) - Change this week: 0 - Registry status: active - Liveness: live - Owner verified: no - Last scored: 2026-11-05 ## Components - remote · `mcp.divux.si`: 73/100 (this document), [markdown](https://verifymcp.io/servers/si-divux-divux/mcp.md), [page](https://verifymcp.io/servers/si-divux-divux/mcp) ## Channel facts - Endpoint: `https://mcp.divux.si` - Transports: `streamable-http` - Auth: `none` - Version: `1.1.0` ## Trust breakdown How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog) Scored 2026-11-05. - **Endpoint Security**: 80/100 - The endpoint's TLS certificate is valid, in date, and uses a strong key. - No authorisation is required to call this server. Every tool declares its destructiveHint and none is destructive, so open access doesn't expose one. - HTTPS is enforced; there's no plaintext access path. - The HSTS (Strict-Transport-Security) header is present. - DNSSEC check failed: this domain isn't protected by DNSSEC. - **Transport & Reachability**: 100/100 - Verified streamable-http transport via a live MCP handshake. - **Schema Quality & AI Usability**: 74/100 - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name). - AI-judged instruction clarity (excellent). - Context-footprint check failed: tool/resource definitions use about 7777 tokens (~324/item across 24 items; 24 tools + 0 resources), over budget; trim descriptions and params. - Usage-examples check failed: none of the tools include examples. - **Stability & Change Management**: 1/100 - Stability check failed: schema churn in the 11 days we've observed: 0 tool removals, 4 breaking changes, 0 auth/transport breaks, 18 additions. - **Tool Coverage**: 99/100 - 100% of tools have a non-trivial description (not blank, and not just the tool's name). - 96% of tool parameters carry a description. - **Tool Safety**: 100/100 - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured. - All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation. - An AI judge read all 25 captured unit(s) of tool text and found none that tries to manipulate the model reading it. - **Capabilities**: 100/100 - Implements a current MCP spec version (2026-07-28). ## Install ### How do I install the divux MCP server? divux is a hosted endpoint at https://mcp.divux.si/, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation. ### Claude ```bash claude mcp add --transport http si-divux-divux 'https://mcp.divux.si/' ``` ### Cursor ```json { "mcpServers": { "si-divux-divux": { "url": "https://mcp.divux.si/" } } } ``` ### VS Code ```json { "servers": { "si-divux-divux": { "type": "http", "url": "https://mcp.divux.si/" } } } ``` ### Codex ```toml [mcp_servers.si-divux-divux] url = "https://mcp.divux.si/" ``` ### opencode ```json { "$schema": "https://opencode.ai/config.json", "mcp": { "si-divux-divux": { "type": "remote", "url": "https://mcp.divux.si/", "enabled": true } } } ``` ### OpenClaw ```bash openclaw mcp add si-divux-divux --url 'https://mcp.divux.si/' --transport streamable-http ``` ### Hermes ```yaml mcp_servers: si-divux-divux: url: "https://mcp.divux.si/" ``` ### Netclaw ```json { "McpServers": { "si-divux-divux": { "Transport": "http", "Url": "https://mcp.divux.si/" } } } ``` ### Vellum ```bash assistant mcp add si-divux-divux -t streamable-http -u 'https://mcp.divux.si/' ``` ### Other ```json { "mcpServers": { "si-divux-divux": { "type": "http", "url": "https://mcp.divux.si/" } } } ``` The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs. ## Changelog Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims. ### 2026-11-04 (score 73, 0) - [security regression] A breaking change shipped without a version bump: still 0.1.0 - [security regression] Tool “game_deals” was removed - [security] Tool “check_illness” rewrote its description, which is the text the model reads - [security] Tool “check_place” rewrote its description, which is the text the model reads - [security] Tool “check_trials” rewrote its description, which is the text the model reads - [security] Tool “check_trip” rewrote its description, which is the text the model reads - [security] Tool “plan_recipe” rewrote its description, which is the text the model reads - [cosmetic] “check_place” added an optional parameter “location” - [cosmetic] “check_trip” added an optional parameter “city” - [cosmetic] “check_place” reworded the description of “compare_to” - [cosmetic] “check_place” reworded the description of “schools” - [cosmetic] “check_trials” reworded the description of “place” - [cosmetic] “check_place” dropped the optional parameter “address” - [cosmetic] “check_place” dropped the optional parameter “lat” - [cosmetic] “check_place” dropped the optional parameter “lon” - [cosmetic] “check_trials” dropped the optional parameter “lat” - [cosmetic] “check_trials” dropped the optional parameter “lon” - [cosmetic] “check_trip” dropped the optional parameter “lat” - [cosmetic] “check_trip” dropped the optional parameter “lon” ### 2026-11-02 (score 73, 0) - [security] The server rewrote its instructions, which are the text every model session reads - [security] Tool “check_car” rewrote its description, which is the text the model reads - [security] Tool “check_care” rewrote its description, which is the text the model reads - [security] Tool “check_company” rewrote its description, which is the text the model reads - [security] Tool “check_cosmetic” rewrote its description, which is the text the model reads - [security] Tool “check_food” rewrote its description, which is the text the model reads - [security] Tool “check_illness” rewrote its description, which is the text the model reads - [security] Tool “check_inflation” rewrote its description, which is the text the model reads - [security] Tool “check_kids” rewrote its description, which is the text the model reads - [security] Tool “check_kitchen” rewrote its description, which is the text the model reads - [security] Tool “check_medicine” rewrote its description, which is the text the model reads - [security] Tool “check_pet” rewrote its description, which is the text the model reads - [security] Tool “check_place” rewrote its description, which is the text the model reads - [security] Tool “check_rates” rewrote its description, which is the text the model reads - [security] Tool “check_recalls” rewrote its description, which is the text the model reads - [security] Tool “check_trials” rewrote its description, which is the text the model reads - [security] Tool “check_trip” rewrote its description, which is the text the model reads - [security] Tool “check_wage” rewrote its description, which is the text the model reads - [security] Tool “compare_prices” rewrote its description, which is the text the model reads - [security] Tool “compare_transfers” rewrote its description, which is the text the model reads - [security] Tool “find_energy_star” rewrote its description, which is the text the model reads - [security] Tool “game_deals” rewrote its description, which is the text the model reads - [security] Tool “plan_build” rewrote its description, which is the text the model reads - [security] Tool “plan_recipe” rewrote its description, which is the text the model reads - [security] Tool “search_products” rewrote its description, which is the text the model reads - [security] Tool “send_feedback” rewrote its description, which is the text the model reads - [cosmetic] “check_food” reworded the description of “items” - [cosmetic] “check_kids” reworded the description of “country” - [cosmetic] “check_medicine” reworded the description of “country” - [cosmetic] “check_medicine” reworded the description of “with” - [cosmetic] “check_recalls” reworded the description of “country” - [cosmetic] “check_trip” reworded the description of “from” - [cosmetic] “check_trip” reworded the description of “legs” - [cosmetic] “check_trip” reworded the description of “to” - [cosmetic] “compare_prices” reworded the description of “condition” - [cosmetic] “compare_prices” reworded the description of “detail” - [cosmetic] “compare_prices” reworded the description of “product_id” - [cosmetic] “compare_transfers” reworded the description of “quotes” - [cosmetic] “plan_build” reworded the description of “parts” - [cosmetic] “plan_build” reworded the description of “picket_planter” - [cosmetic] “plan_build” reworded the description of “stock” - [cosmetic] “plan_recipe” reworded the description of “days” - [cosmetic] “plan_recipe” reworded the description of “recipe” - [cosmetic] “search_products” reworded the description of “brand” - [cosmetic] “search_products” reworded the description of “condition” - [cosmetic] “search_products” reworded the description of “max_price” - [cosmetic] “search_products” reworded the description of “min_price” - [cosmetic] “search_products” reworded the description of “query” - [cosmetic] “send_feedback” reworded the description of “details” - [cosmetic] “send_feedback” reworded the description of “kind” - [cosmetic] “send_feedback” reworded the description of “summary” - [cosmetic] “send_feedback” reworded the description of “tool” - [cosmetic] Tool “check_car” changed its title: Car: official facts about a vehicle, and fuel cost → Car: official vehicle data and fuel cost - [cosmetic] Tool “check_company” changed its title: Company: is a company real, its complaints and its SEC filings → Company: registry record, complaints and SEC filings - [cosmetic] Tool “check_kids” changed its title: Kids: recalls and safety for baby and child gear → Kids: recalls for baby and child gear - [cosmetic] Tool “check_pet” changed its title: Pets: pet food recalls and what is toxic to dogs and cats → Pets: pet food recalls and a list of foods and plants toxic to dogs and cats - [cosmetic] Tool “check_trials” changed its title: Trials: clinical trials recruiting near you → Trials: recruiting clinical trials for a condition - [cosmetic] Tool “check_wage” changed its title: Wage: what a job pays where you live → Wage: what a job pays, by place ### 2026-11-01 (score 73, 0) - [security] Tool “check_car” rewrote its description, which is the text the model reads - [cosmetic] “check_car” added an optional parameter “odometer” - [cosmetic] “check_car” added an optional parameter “zip” ### 2026-10-31 (score 73, 0) - [security] Tool “check_kitchen” rewrote its description, which is the text the model reads - [security] Tool “check_recalls” rewrote its description, which is the text the model reads - [security] Tool “check_trip” rewrote its description, which is the text the model reads - [security] Tool “plan_build” rewrote its description, which is the text the model reads - [security] Tool “plan_recipe” rewrote its description, which is the text the model reads - [cosmetic] “check_trip” added an optional parameter “legs” ### 2026-10-30 (score 73, 0) - [security] Tool “check_kids” rewrote its description, which is the text the model reads - [security] Tool “check_kitchen” rewrote its description, which is the text the model reads - [security] Tool “check_pet” rewrote its description, which is the text the model reads - [security] Tool “check_place” rewrote its description, which is the text the model reads - [security] Tool “check_recalls” rewrote its description, which is the text the model reads - [security] Tool “check_trip” rewrote its description, which is the text the model reads - [security] Tool “compare_prices” rewrote its description, which is the text the model reads - [security] Tool “plan_recipe” rewrote its description, which is the text the model reads ### 2026-10-29 (score 73, 0) - [security] Tool “check_wage” rewrote its description, which is the text the model reads - [security] Tool “plan_build” rewrote its description, which is the text the model reads - [functional regression] Schema quality: 303 → 357 - [functional] Destructive annotations: pass → 100 - [functional] New tool “check_kids” - [functional] New tool “check_kitchen” - [functional] New tool “check_pet” - [functional] New tool “plan_recipe” - [functional] New tool “send_feedback” - [cosmetic] “check_place” added an optional parameter “compare_to” - [cosmetic] “plan_build” added an optional parameter “picket_planter” - [cosmetic] “check_trials” reworded the description of “place” - [cosmetic] “plan_build” reworded the description of “parts” - [cosmetic] “plan_build” made “parts” optional ### 2026-10-28 (score 73, +7) - [security improvement] HSTS header: fail → pass - [security improvement] HTTPS: fail → pass - [security] Tool “check_trip” rewrote its description, which is the text the model reads - [security] Tool “game_deals” rewrote its description, which is the text the model reads - [security] Tool “plan_build” rewrote its description, which is the text the model reads - [cosmetic] “plan_build” added an optional parameter “units” - [cosmetic] “plan_build” reworded the description of “kerf” - [cosmetic] Tool “check_car” changed its title: Car check: official facts about a vehicle, and fuel cost → Car: official facts about a vehicle, and fuel cost - [cosmetic] Tool “check_care” changed its title: Care check: doctors, hospitals and nursing homes → Care: doctors, hospitals and nursing homes - [cosmetic] Tool “check_company” changed its title: Company check: is a company real, its complaints and its SEC filings → Company: is a company real, its complaints and its SEC filings - [cosmetic] Tool “check_cosmetic” changed its title: Cosmetics check: what's in skincare and makeup, and EU rules → Cosmetics: what's in skincare and makeup, and EU rules - [cosmetic] Tool “check_food” changed its title: Food check: nutrition, ingredients and recalls → Food: nutrition, ingredients and recalls - [cosmetic] Tool “check_illness” changed its title: Illness check: what's going around → Illness: what's going around - [cosmetic] Tool “check_medicine” changed its title: Medicine check: FDA label, recalls, shortages, pharmacy cost → Medicine: FDA label, recalls, shortages, pharmacy cost - [cosmetic] Tool “check_place” changed its title: Place check: hazards, weather, schools, parks and the night sky → Place: hazards, weather, schools, parks and the night sky - [cosmetic] Tool “check_rates” changed its title: Rates check: interest rates now, after inflation → Rates: interest rates now, after inflation - [cosmetic] Tool “check_trials” changed its title: Trials check: clinical trials recruiting near you → Trials: clinical trials recruiting near you - [cosmetic] Tool “check_wage” changed its title: Wage check: what a job pays where you live → Wage: what a job pays where you live ### 2026-10-27 (score 66, −1) - [security regression] Stability: 0.03 → fail - [security regression] A breaking change shipped without a version bump: still 0.1.0 - [security regression] Tool “check_park” was removed - [security regression] Tool “check_schools” was removed - [security regression] Tool “check_sky” was removed - [security] The server rewrote its instructions, which are the text every model session reads - [security] Tool “check_place” rewrote its description, which is the text the model reads - [functional regression] Tool “check_recalls” dropped its output schema - [functional regression] Tool “compare_prices” dropped its output schema - [functional regression] Tool “find_energy_star” dropped its output schema - [functional regression] Tool “search_products” dropped its output schema - [functional improvement] Schema quality: 357 → 295 - [cosmetic] “check_car” added an optional parameter “fuel_only” - [cosmetic] “check_place” added an optional parameter “college” - [cosmetic] “check_place” added an optional parameter “level” - [cosmetic] “check_place” added an optional parameter “limit” - [cosmetic] “check_place” added an optional parameter “major” - [cosmetic] “check_place” added an optional parameter “park” - [cosmetic] “check_place” added an optional parameter “place” - [cosmetic] “check_place” added an optional parameter “radius_miles” - [cosmetic] “check_place” added an optional parameter “schools” - [cosmetic] “check_place” added an optional parameter “sky” - [cosmetic] “check_place” added an optional parameter “state” - [cosmetic] Tool “check_place” changed its title: Place check: flood zone, hazard risk and weather alerts → Place check: hazards, weather, schools, parks and the night sky ## MCP tools (24) ### `search_products` (~265 tokens) Shopper: search products Returns matching products and their ids for a name, model number, barcode (UPC/EAN), or a brand plus category such as 'eufy security cameras'. One call returns up to 10 products, with the stores that list each one and a link to each product's Divux page. Prices and stock are for online delivery in the US. An empty result means no product matched the query. Recall notices are not part of this result. Applies to questions about finding a specific product, or a brand's range, to buy in the US. Input parameters: - `brand` (string): Brand filter, e.g. 'Dyson' - `condition` (string): new or used. 'used' includes open-box and refurbished. Default is new. - `in_stock_only` (boolean): Only products with at least one in-stock offer - `limit` (integer) - `max_price` (number): Highest price in USD, e.g. 100 - `min_price` (number): Lowest price in USD - `query` (string, required): A product name, model number or UPC/EAN, e.g. 'Sony WH-1000XM6' or '045496884963', or a brand plus category, e.g. 'eufy security cameras' ### `compare_prices` (~206 tokens) Shopper: compare prices Returns current US store offers for one Divux product id, ranked by price (item plus shipping where the store lists it, tax excluded). The default is the top 3 offers; detail set to true returns up to 8. Each offer includes a link to its Divux page. Official CPSC recall notices that mention the product are included when some exist; their absence does not mean a product has no recall. Links may be paid links; the order is by price only. Divux does not check out. An empty offer list means no matching offer was found. Applies to questions about current prices and stores for one specific product. Input parameters: - `condition` (string): new, used or any. 'used' includes open-box and refurbished. Default is new. - `detail` (boolean): false returns the top 3 offers. true returns up to 8. - `include_out_of_stock` (boolean) - `product_id` (string, required): Product id from a product search ### `plan_build` (~530 tokens) Builder: plan a build Returns a cut list, a buy list and a plan page with a drawing, cut diagrams and a 3D view for a project given as pieces (name, material, size in inches, position 'at') or as a cedar fence-picket planter (picket_planter: length, width, courses, leg_below). It also returns amounts to buy for materials given by area, volume or count, such as paint, soil, concrete or tile (quantities). Nominal lumber sizes are converted to actual sizes (a 2x4 is 1.5 by 3.5 in). Omitted board lengths, wood type or sheet sizes are filled with common US store sizes and listed under Assumed. The picket planter design includes corner legs, ledger strips under the floor, a center brace when a side is over 4 ft, a ripped floor filler and trim. Prices appear only when stock, extras or quantities carry a price estimate; store prices are not looked up. The saw kerf defaults to 1/8 inch. Divux saves each plan on a page at a Divux link, up to 25 plans a day per connection. Applies to questions about working out a woodworking or building project, or a materials quantity, into cuts and a shopping list. Input parameters: - `extras` (array): Screws, nails, glue, finish, and tools needed - `kerf` (number): Saw blade width in inches; if left out, divux assumes 1/8 inch and says so under Assumed - `parts` (array): Piece list for projects other than a cedar picket planter. When picket_planter is also present, picket_planter is used. - `picket_planter` (object): Cedar fence-picket planter dimensions. When set, Divux designs the pieces. For any other project, parts holds the pieces. - `quantities` (array): Materials bought by area, volume or count: paint, stain, tile, flooring, fabric, soil, mulch, gravel, concrete - `steps` (array): Assembly steps in order; pieces refer to them with 'step' - `stock` (array): Board lengths or sheets to buy, with a price estimate in USD for one. When omitted, common US store lengths are used and no prices are shown. - `title` (string, required): Project name, e.g. 'Cedar planter 24 x 12 in' - `units` (string): Units for the reply text; imperial (inches) if left out. The plan page has its own switch. ### `plan_recipe` (~464 tokens) Recipes: nutrition, allergens and a shopping list Returns source-linked food chemistry explanations for a small set of recognized ingredient combinations, when present, and a page with nutrition per serving and per day (USDA FoodData Central), allergen flags from ingredient names, any active FDA recall on a named brand, the recipe scaled to the requested servings, and one shopping list that merges the same ingredient across the week and rounds up to amounts shops sell, with price links. Inputs are one recipe or up to a week of days (each meal has servings cooked and servings one person eats). Amounts are a number plus a unit (g, kg, oz, lb, ml, l, tsp, tbsp, cup, or a count); grams may be supplied for a can or a branded item. A brand is recorded only when one is supplied. Optional household details (adults, children's ages, allergies, foods to avoid) scale portions and flag conflicts; a weekly budget with a price estimate on each ingredient totals the list. Divux saves the result on a page at a Divux link, up to 25 plans a day per connection. Allergen and diet flags are based on ingredient names; the package label remains the reference. The numbers are nutrition data, not a statement that a meal is healthy or suits a medical condition. Not medical or diet advice. Applies to questions about the nutrition, allergens and shopping list for a recipe or a week of meals. Input parameters: - `budget` (number): Weekly food budget in USD. With price_estimate on ingredients, Divux totals the shopping list and warns when it's over. - `days` (array): A week of meals, up to 21 recipes in all. Omitted when one recipe is set. - `diet` (array): Diets to check the ingredients against - `goals` (object): The person's own targets per day; the page's bars are drawn against them instead of FDA Daily Values - `household` (object): Family meals: who eats. Divux scales portions per person, and flags or rejects ingredients that break an allergy or avoid. - `recipe`: One recipe. Omitted when days is set. - `title` (string): Name for a week plan, e.g. 'High-protein week'. A single recipe uses its own title. ### `check_trip` (~528 tokens) Trip check: destination advice, fares and airport delays Returns, for the inputs that are present: destination advice for a country (US State Department, UK FCDO and Canada, side by side), active disaster alerts, weather for the city being visited, CDC travel health notices, public holidays between two dates, and an exchange rate; the cheapest round-trip or one-way fares recently found for a route (dates, airline, stops, booking links), which are cached finds and not live quotes; a per-leg total for up to four planned legs, in USD and a local currency; and FAA delay reports for US airport codes (ground stops, delay programs, arrival and departure delays, closures). Booking links may be paid links. Divux does not book flights. Applies to questions about travel to a country, flight prices on a route, or current delays at US airports. Results report what the official sources say and are not a safety assessment of a destination. Input parameters: - `airports` (array): US airport codes for the FAA's delays right now, e.g. ['SFO', 'JFK']; an empty list for every US airport with a delay - `city` (string): City being visited, for weather, e.g. 'Kyoto' or 'Lisbon'. - `country` (string): Destination country in English or two-letter code, e.g. 'Japan' or 'JP' - `depart` (string): For fares: departure month YYYY-MM or date YYYY-MM-DD; defaults to start_date - `direct_only` (boolean): For fares: true for nonstop flights only - `end_date` (string): Trip end, YYYY-MM-DD - `from` (string): For fares: departure city, airport code, or metro, e.g. 'New York', 'JFK', or 'Bay Area' - `home_currency` (string): Traveler's currency, e.g. 'USD' - `legs` (array): Up to four flight legs. The result is cached cheapest fares per leg, by day, and a trip total. - `local_currency` (string): Destination currency, e.g. 'JPY' - `one_way` (boolean): For fares: true for one-way; defaults to one-way when there's no return - `return` (string): For fares: return month or date; defaults to end_date; leave both out for one-way - `start_date` (string): Trip start, YYYY-MM-DD - `to` (string): For fares: destination city, airport code, or metro, e.g. 'Lisbon' or 'Tokyo'; defaults to the country's capital ### `check_car` (~453 tokens) Car: official vehicle data and fuel cost Returns official US vehicle data for a VIN or a year, make and model: specs from NHTSA vPIC, NHTSA recalls, owner complaints with the most-reported components, NHTSA 5-star crash ratings, and EPA fuel economy from fueleconomy.gov. With miles, it also returns fuel cost: EPA combined mpg (or a supplied mpg) times the current weekly average pump price from the US Energy Information Administration for a state, its region or the US. With a VIN, odometer and ZIP code, it also returns a market price for that car, plus the median and range of similar cars listed for sale. A VIN selects one version; a year, make and model covers the model year. Accident history and title status are not included. The value is a model estimate, not an appraisal. Applies to questions about a specific vehicle, or the fuel cost of a drive. Results report what the official sources say and are not a safety assessment. Input parameters: - `fuel` (string): For fuel cost, when the model comes in both - `fuel_only` (boolean): true when only fuel cost is wanted; skips recall, complaint and crash-rating lookups (needs miles and mpg, or year/make/model) - `make` (string): Make, e.g. 'Honda' - `miles` (number): For fuel cost: driving distance in miles, one way - `model` (string): Model, e.g. 'CR-V' - `mpg` (number): For fuel cost: the car's mpg, if known or when there's no specific car - `odometer` (number): For fair value: the car's current mileage on the odometer (not a trip distance) - `round_trip` (boolean): For fuel cost: true to double the distance - `state` (string): For fuel cost: two-letter US state where most of the driving happens, e.g. CA - `vin` (string): 17-character VIN, if known - `year` (integer): Model year, when there's no VIN - `zip` (string): For fair value: the 5-digit US ZIP code where the car is or will be sold ### `compare_transfers` (~231 tokens) Money abroad: compare transfer quotes Returns the amount that arrives for each supplied provider quote, the exchange-rate markup against the ECB reference rate, the total cost as a percent, a ranking, and the World Bank average cost for the receiving country. Inputs are the amount, the two currency codes, and each provider's fee and rate (receiving currency per one unit of sending currency). A rate supplied the other way round is flipped and the flip is noted. With no quotes, the result is the reference rate and the benchmark. Applies to questions about comparing providers for sending money to another country. Not financial advice; Divux does not send money. Input parameters: - `quotes` (array): One quote per provider: name, fee in the sending currency, and rate. - `receive_country` (string): Receiving country's two-letter code, e.g. 'MX', for the World Bank benchmark - `receive_currency` (string, required): Receiving currency, e.g. 'MXN' - `send_amount` (number, required): Amount being converted, before fees, in the sending currency - `send_currency` (string, required): Sending currency, e.g. 'USD' ### `check_food` (~336 tokens) Food: nutrition, ingredients and recalls Returns nutrition for a food from its barcode (UPC/EAN, via Open Food Facts) or its name (USDA FoodData Central): calories, protein, fat, saturated fat, carbs, sugars, fiber and sodium per 100 g or 100 ml, per serving and for gram amounts when the source uses grams, with % Daily Value from the FDA's reference values. Volume-based foods are not converted to grams without a density. Barcodes also return ingredients, allergens and Nutri-Score. When a brand is known, FDA recall reports matching it are included. A list of items (name or barcode with grams or servings, or menu items with per-serving numbers) plus optional daily targets returns the sum, % Daily Value and what remains of each target. Applies to questions about the nutrition, ingredients or recalls of a food, or the total of a meal. Results are nutrition numbers and are not a statement that a food is healthy or safe. Not medical or diet advice. Medicines and cosmetics are outside this result. Input parameters: - `amount_g` (number): Amount eaten in grams, to scale the values - `barcode` (string): UPC or EAN printed under the bars, 8 to 14 digits - `fdc_id` (integer): A USDA FoodData Central id from an earlier result's alternatives - `items` (array): Meal items. The result is their sum. - `query` (string): Food name, with brand when known, e.g. 'Chobani plain Greek yogurt' or 'banana' - `targets` (object): The person's own daily or meal targets, for what's left ### `check_cosmetic` (~271 tokens) Cosmetics: what's in skincare and makeup, and EU rules Returns what the EU Cosmetics Regulation says about the ingredients of a skincare, makeup, hair or body product, from the European Commission's CosIng annexes: banned (Annex II, with any exception), restricted with the maximum concentration, where it may be used and the required warnings (Annex III), fragrance allergens the EU requires on the label, and preservatives, UV filters and colorants allowed only within limits. Inputs are an INCI ingredient list, or a product name with brand or a barcode looked up in Open Beauty Facts. Recalls and EU Safety Gate alerts naming the brand (EU, UK and Canada) are included. The result does not rate products or say what suits a person's skin. Applies to questions about what the EU rules say about the ingredients of a skincare, makeup, hair or body product. Results report what the official sources say and are not a safety assessment. Not medical advice. Medicines and foods are outside this result. Input parameters: - `barcode` (string): UPC or EAN on the package - `ingredients` (string): The INCI ingredient list, comma separated, e.g. 'Aqua, Glycerin, Niacinamide, Phenoxyethanol' - `product` (string): Product name with brand, e.g. 'CeraVe Moisturizing Cream' ### `check_pet` (~252 tokens) Pets: pet food recalls and a list of foods and plants toxic to dogs and cats Returns FDA animal and veterinary recall notices for a pet food or treat brand or product, newest first, with the reason, the company and the notice link. Also returns a curated list entry for common foods and plants (chocolate, grapes and raisins, xylitol, onions and garlic, macadamia nuts, alcohol, salt, lilies for cats, sago palm, azalea, oleander, tulips, daffodils and more): the source's wording, its page, the date it was read, and poison-control phone numbers (fees may apply). A name that is not on the list has no entry. Applies to questions about pet food recalls, or whether a food or plant is on the list of items toxic to dogs or cats. Results report what the official sources say and are not a safety assessment. Not veterinary advice. Human-food nutrition is outside this result. Input parameters: - `query` (string, required): A pet food or treat brand or product (e.g. 'Fromm', 'Pedigree wet dog food') or a food or plant name (e.g. 'grapes', 'lily', 'xylitol') - `species` (string): The pet, when known ### `check_kitchen` (~578 tokens) Kitchen: cooking temperatures, storage times, baking math, ratios and swaps Returns kitchen figures in F and C: the federal minimum internal temperature for poultry, ground meat, beef, bison, goat, pork, lamb, veal, ham, rabbit, venison, fish, egg dishes and leftovers with rest times, and a shellfish visual check that has no number; fridge and freezer keeping times for about 40 foods; the curing-salt limit (the federal 156 ppm figure only); the high-altitude boiling point. Trade-practice answers are labelled as trade practice: brine and quick-pickle ratios, coffee and tea ratios, baker's percentages, recipe scaling, and ingredient swaps with allergy flags and a one-line why. Topic is temperature, storage, baking, ratio or substitute, and is inferred from the query when topic is omitted. Applies to questions about minimum internal cooking temperatures, how long food keeps, baking math, brine, coffee or tea ratios, or an ingredient swap. The figures are cooking guidance, not a nutrition panel or a recall notice. Results report what the official sources say and are not a safety assessment. Not medical advice. Input parameters: - `altitude_ft` (number): Altitude in feet - `altitude_m` (number): Altitude in metres - `amount` (number): For ratios: the water or liquid (brine, pickle, coffee, tea) or the meat weight (dry brine, cure) - `avoid` (array): Allergens to avoid; swaps that list them are flagged and sorted last - `factor` (number): Scale factor, e.g. 1.5 - `food` (string): The food, tea type or ingredient, e.g. 'chicken', 'salmon', 'buttermilk', 'green tea' - `from_servings` (number): Servings the recipe makes - `ingredients` (array): For baking scale or percent: the recipe's ingredients with amounts. Weigh flour and water in g, oz or lb for percentages - `kind` (string): For baking: scale, percent (baker's percentages and hydration) or altitude. For ratio: wet_brine, dry_brine, pickle, cure, coffee or tea - `query` (string): The question in the person's words, e.g. 'how long does cooked rice keep' - `reading_c` (number): A thermometer reading in C to compare with the food's minimum - `reading_f` (number): A thermometer reading in F to compare with the food's minimum - `to_servings` (number): Servings wanted - `topic` (string): Which kind of answer; inferred from query or the other fields when left out - `unit` (string): Unit for amount: g, kg, oz, lb, ml, l, tsp, tbsp, cup or floz ### `send_feedback` (~207 tokens) Report a problem or idea to Divux Stores a short report for Divux: a wrong dated figure, a tool that failed, a tool that is missing, or an idea. Fields are kind, a summary of up to 300 characters, an optional tool name, and optional details of up to 1,000 characters. The text is stored as sent, so the fields are for general descriptions without personal details. Divux keeps the text and does not reply. The limit is 5 reports a day per connection. Applies to questions about reporting a problem with Divux or suggesting an idea to Divux. Input parameters: - `details` (string): A short general description of the problem, without personal details. - `kind` (string, required): Type of report: a wrong dated figure, a broken tool, a missing tool or an idea. - `summary` (string, required): The problem or idea in one or two general sentences, without personal details. - `tool` (string): Name of the Divux tool involved, when one was involved. ### `check_place` (~523 tokens) Place: hazards, weather, schools, parks and the night sky Returns, for a US city and state or ZIP code (point checks such as the flood zone are for the area's center, not one home): flood zone, FEMA National Risk Index ratings, National Weather Service alerts, air quality, crime rate, drinking water, internet providers, electricity price, home-price trends, EV chargers, county health, childcare prices and nearby HUD housing counselors. A point outside the US returns earthquakes, GDACS disaster alerts within 500 km and the 10-day weather risk. Optional parts, when those inputs are present: a US national park (alerts, closures, fees, campgrounds); night sky at an area's center (aurora chance, moon, darkness, tides); public schools nearby (not school assignment or test scores); a US college (net price after aid, graduation rate, debt, earnings, and a major when supplied). compare_to sets a second US city or ZIP code beside the first and lists the numbers that differ. Applies to questions about conditions and services in a US city or ZIP code, a national park, the night sky, nearby schools or a US college. Results report what the official sources say and are not a safety assessment or an insurance quote. Input parameters: - `college` (string): A US college's name, e.g. 'University of Washington', 'Rice University' - `compare_to` (string): A second US city and state or ZIP code, to set next to `location` and list the numbers that differ. Not a ranking. - `level` (string): School level, with schools - `limit` (integer): How many schools - `location` (string): A US city and state or a ZIP code, e.g. 'Austin, TX' or '78701'. - `major` (string): A field of study at the college, for what its graduates earn and borrow, e.g. 'nursing' - `park` (string): A US national park or site's name or 4-letter NPS code, e.g. 'Yosemite', 'mora' - `place` (string): Name of the place, for the answer - `radius_miles` (number): School search radius, default 2 - `schools` (boolean): true for public schools near the area center, closest first - `sky` (boolean): true for the night sky at the point: aurora chance, moon, darkness and, on US coasts, tides - `state` (string): Two-letter state to narrow a common college name, e.g. 'WA' ### `check_medicine` (~281 tokens) Medicine: FDA label, recalls, shortages, pharmacy cost Returns official records for a medicine from its brand or generic name (e.g. 'Advil', 'metformin'). For an EU or EEA country: the European Medicines Agency's authorisation, what it is approved for, shortages and safety letters. For the US (the default): whether it is over the counter or prescription, what the FDA label says it is for, its boxed warning and key warnings, a link to the full label (DailyMed), FDA recall reports, current FDA drug shortages, and what US pharmacies pay per tablet for each strength (CMS NADAC), brand next to generic. With `with` set to a second medicine, the result includes the label sentences where either label names the other; the absence of a mention is not evidence about safety. Applies to questions about what official records say about a medicine: its label, recalls, shortages or pharmacy cost. Not medical advice. Foods and medical devices are outside this result. Input parameters: - `country` (string): Country whose regulator is used: an EU or EEA country uses the European Medicines Agency; the US (the default) uses the FDA - `name` (string, required): Brand or generic name, e.g. 'Advil' or 'metformin' - `with` (string): Optional second medicine name; the result lists label sentences in either label that name the other ### `check_inflation` (~217 tokens) Money value: inflation math from the official CPI Returns inflation math for a country: what an amount from one year is worth in another, the total price change, the average yearly rate and the latest inflation rate. The US series is the Bureau of Labor Statistics CPI-U, monthly, from 1913. Other countries use the national consumer price index as published by the World Bank, yearly. Inputs are amount, from, to and country. Applies to questions about what an amount from one year is worth in another year, or how prices changed over time. Today's interest rates and pay for a job are outside this result. Input parameters: - `amount` (number, required): Amount in the country's currency (dollars for the US) - `country` (string): Country name or two-letter code; defaults to the United States. Amounts are in that country's currency. - `from` (string, required): Year (1990), or for the US a month (1990-06), the amount is from - `to` (string): Year or month to convert to; defaults to the latest figure ### `check_care` (~345 tokens) Care: doctors, hospitals and nursing homes Returns clinicians from the federal NPI registry (CMS) for a name (last name, first name, state) or a specialty in a city or state: credential, specialty, state license number, practice address and phone, whether the record is active, and whether they opted out of Medicare. The registry is not a quality rating and does not show insurance networks. For hospitals and nursing homes, returns Medicare's ratings in a US city or ZIP code, highest rated first. Nursing homes include overall, health-inspection, staffing and quality stars, total fines, Medicare's abuse icon and stale inspections. Hospitals include the overall star rating, emergency services and safety measures better or worse than national. Each row includes a Medicare Care Compare link. Applies to questions about a clinician in the federal registry, or Medicare ratings for hospitals and nursing homes in a US place. Not medical advice; the person chooses their care. Input parameters: - `city` (string) - `first_name` (string): Clinician's first name - `kind` (string): clinician (doctor, nurse practitioner, therapist and others), hospital or nursing_home; clinician when a name or specialty is sent - `last_name` (string): Clinician's last name - `limit` (integer) - `name` (string): Part of a hospital's or nursing home's name, to check one place - `specialty` (string): Clinician specialty, e.g. 'pediatrics', 'dermatology', 'family medicine', 'psychologist' - `state` (string): Two-letter state - `zip` (string): Hospitals and nursing homes: 5-digit ZIP code instead of city and state ### `check_company` (~239 tokens) Company: registry record, complaints and SEC filings Returns, for a company outside the US, whether it is in the global LEI register (GLEIF): legal name, country, local registry number, whether it is active and its ultimate parent. For a US company, returns the Consumer Financial Protection Bureau complaint record (counts, top problems, on-time answers, how often people got relief) and whether it is an FDIC-insured bank, and for a public company what it reported to the SEC (EDGAR XBRL): revenue, net income and diluted EPS for four fiscal years, the latest quarter against a year earlier, cash and shares outstanding. The input is a name or a US ticker such as NVDA. Complaint counts grow with company size. Applies to questions about what public records show about a company. Not investment advice. Stock prices are not included. Input parameters: - `company` (string, required): Company name or US stock ticker, e.g. 'Capital One', 'Chime', 'NVDA' - `country` (string): Country where the company is registered or operates; leave out for the US - `months` (integer): How far back to count; default 24 ### `check_trials` (~209 tokens) Trials: recruiting clinical trials for a condition Returns clinical trials recruiting now for a condition, from ClinicalTrials.gov: title, phase, sponsor, who can join (age, sex, healthy volunteers), the nearest recruiting site and its distance, and the study page. A city and state or ZIP code searches within a radius. An age keeps only trials that accept that age. Applies to questions about which clinical trials are recruiting for a condition. Not medical advice. A medicine label, a recall report and a pharmacy price are outside this result. Input parameters: - `age` (integer) - `condition` (string, required): Condition, e.g. 'type 1 diabetes', 'early-stage triple negative breast cancer' - `intervention` (string): A drug or treatment to narrow to, e.g. 'pembrolizumab' - `limit` (integer) - `place` (string): A city and state or ZIP code, e.g. 'Boston, MA' or '02108'. - `radius_miles` (number): Default 50 ### `check_rates` (~193 tokens) Rates: interest rates now, after inflation Returns the central bank's policy rate for a country (Bank for International Settlements, about 40 central banks, the ECB for euro countries), the date it last changed, and the country's latest yearly inflation. For the US, returns Treasury yields from the Treasury's daily par yield curve (1 month to 30 years), each next to the latest 12-month inflation (BLS CPI) as a yield minus trailing CPI, and optionally the interest an amount would earn at one term. Applies to questions about current central-bank policy rates, or US Treasury yields next to inflation. Not investment advice. The change in purchasing power of money from a past year is outside this result. Input parameters: - `amount` (number): Dollar amount to show interest for - `country` (string): Country name or two-letter code; defaults to the United States - `term` (string): Term for the interest example; default 1 year ### `check_illness` (~238 tokens) Illness: what's going around Searches up to the latest 100 World Health Organization Disease Outbreak News reports for a country in the past 12 months, and discloses when the available reports cover a shorter period. For a US state, returns the latest available dated CDC figures: the state's acute respiratory illness level, wastewater levels for COVID, flu A and RSV at a typical site with the direction over three weeks (and the sites serving a county when one is given), and whether measles or H5 bird flu was found in the state's wastewater in the last four weeks. Inputs are a US state and an optional county, or a country. Applies to questions about current illness levels in a US state or reported outbreaks in a country. These are community levels, not a person's risk. Not medical advice. Travel-health notices for a destination are not in this result. Input parameters: - `country` (string): Any other country, for WHO outbreak reports; leave out for the US - `county` (string): US county name without 'County', e.g. 'King' - `state` (string): US state name or two-letter code, e.g. 'WA' ### `check_wage` (~245 tokens) Wage: what a job pays, by place Returns what a job pays. In the US, for a metro area, a state or the whole country, from the Bureau of Labor Statistics Occupational Employment and Wage Statistics (May 2025, all industries): median yearly pay, the 10th to 90th percentile range and how many people do the job. In Canada, from Statistics Canada: median and average hourly pay for the job group, by province. Elsewhere, average monthly pay of employees by broad job group (managers, professionals, technicians, trades and more) in local currency, from the International Labour Organization's ILOSTAT. Inputs are a job title and, in the US, an optional city or state. Applies to questions about what a job pays. The change in purchasing power of old dollars is outside this result. Input parameters: - `country` (string): Country name or two-letter code; defaults to the United States - `job` (string, required): Job title, e.g. 'registered nurse', 'electrician', 'software developer' - `place` (string): US only: city ('Seattle', 'Austin, TX'), state ('WA' or 'Washington'), or leave out for national ### `check_recalls` (~368 tokens) Shopper: check recalls Returns official recall notices. Coverage: US, UK, Canada and EU. In the US, category 'product' searches CPSC consumer-product recalls by words in the product name (brand plus product type, e.g. 'Rowenta vacuum'); 'food_drug' searches FDA food, drug and medical device enforcement reports (e.g. 'CPAP', 'insulin pump'); 'vehicle' searches NHTSA by make, model and model year. country 'GB' searches UK OPSS product recalls and Food Standards Agency alerts; 'CA' Health Canada; an EU or EEA country (e.g. 'DE', 'France') the EU Safety Gate alerts on dangerous non-food products and vehicles from the last three years. An empty result means no notice was found; it does not mean an item is free of recalls elsewhere. Applies to questions about whether a product, food, drug, medical device or vehicle has a recall notice. Results report what the official sources say and are not a safety assessment. Not medical advice. Input parameters: - `category` (string, required): product (CPSC), food_drug (FDA food, drugs and medical devices) or vehicle (NHTSA) - `country` (string): Country whose notices are searched: United States (default), United Kingdom, Canada or any EU/EEA country - `make` (string): Vehicle make, e.g. 'Honda' (vehicle only) - `model` (string): Vehicle model, e.g. 'Civic' (vehicle only) - `query` (string): Product, food, drug or medical device words, for product and food_drug, e.g. 'stroller', 'peanut butter' or 'CPAP' - `year` (integer): Vehicle model year (vehicle only) ### `check_kids` (~306 tokens) Kids: recalls for baby and child gear Returns official recall notices for up to 6 baby or child items in one call: cribs, sleepers, strollers, high chairs, toys, furniture (CPSC), infant formula and baby food (FDA), and car seats. Each item comes back with any recalls, what the notice says to do (stop using, refund, repair kit) and a status. An empty result means no notice was found. Car seats are not searched by name, so they come back as 'check model' with the official recall and rating pages. Safe-sleep and furniture tip-over guidance is included only when the items or the child's age match. A single car that is not a car seat, and a product that is not baby or child gear, are outside this list. Applies to questions about whether baby or child items have recall notices. Results report what the official sources say and are not a safety assessment. Not medical advice. Input parameters: - `ages` (string): The children's ages, e.g. 'newborn, 3 years'; used only to decide whether safe-sleep guidance applies - `country` (string): Country whose notices are searched: United States (default), United Kingdom, Canada or an EU/EEA country - `items` (array, required): Up to 6 things, brand plus product, one per entry: 'Fisher-Price Rock n Play', 'Similac formula', 'IKEA Malm dresser', 'Graco 4Ever car seat' ### `find_energy_star` (~185 tokens) Shopper: find ENERGY STAR models Returns whether an appliance model is ENERGY STAR certified, or a list of certified models (most efficient first), from the EPA's official data. Categories: heat pumps, heat pump water heaters, clothes washers, clothes dryers, refrigerators, dishwashers. Heat pump answers include that federal 25C/25D tax credits ended after Dec 31, 2025. Applies to questions about whether an appliance model is ENERGY STAR certified, or which models in a category are certified. A store price and a recall notice are outside this result. Input parameters: - `brand` (string): Brand to filter by, e.g. 'Rheem' - `category` (string, required): Appliance category - `model` (string): Model number or part of it, to check certification - `most_efficient_only` (boolean): Only models meeting ENERGY STAR Most Efficient criteria (not available for water heaters) ## Diagnostics Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/si-divux-divux/mcp#diagnostics ## Score history - 2026-11-05: 73 - 2026-11-04: 73 - 2026-11-03: 73 - 2026-11-02: 73 - 2026-11-01: 73 - 2026-10-31: 73 - 2026-10-30: 73 - 2026-10-29: 73 - 2026-10-28: 73 - 2026-10-27: 66 - 2026-10-26: 67 - 2026-10-25: 64 ## Common questions ### What is the divux MCP server? divux is an MCP server listed in the public MCP registry as si.divux/divux. Official data for AI assistants: medicines, recalls, cosmetics, money, travel, places, prices. This page covers its hosted endpoint (https://mcp.divux.si). ### Is the divux MCP server safe to use? divux scores 73 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm. ### What tools does the divux MCP server expose? divux exposes 24 tools: search_products, compare_prices, plan_build, plan_recipe, check_trip, and 19 more. Their descriptions and schemas cost roughly 7,670 tokens of context every time the server is loaded. ### Does the divux MCP server require authentication? No. We connected to divux without credentials and it answered, so anything it exposes is reachable by anyone who knows the address. ### Is the divux MCP server still maintained? divux is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced. ## Links - Remote endpoint: https://mcp.divux.si/ - Website: https://divux.si/ - Changelog RSS feed: https://verifymcp.io/servers/si-divux-divux/mcp.xml - Changelog JSON feed: https://verifymcp.io/servers/si-divux-divux/mcp.json - HTML version of this page: https://verifymcp.io/servers/si-divux-divux/mcp