mcp-marmiton
mcp-marmiton Server Capabilities
mcp-marmiton is a read-only MCP server that lets you search, read, and rescale French recipes from Marmiton — no API key, account, or configuration required.
Core Features
Search recipes (
search_recipes): Find Marmiton recipes by dish name or ingredient using free-text queries (best in French, e.g.tarte aux pommes,poulet curry coco). Returns results with recipe IDs, titles, URLs, and images. Supports alimit(up to 30) per call; one page per call in line with Marmiton'srobots.txt.Read a full recipe (
get_recipe): Retrieve complete details for one recipe by Marmiton ID or URL, including ingredients, step-by-step instructions, prep/cook/total times, category, author, rating, nutrition info (as published), yield, and an attribution/source link back to Marmiton.Rescale quantities (via
get_recipe): Pass aservingsparameter (up to 500) to intelligently rescale ingredient quantities. Each ingredient comes back with a scaling flag:scaled— cleanly multiplied (e.g. grams, millilitres)rounded— multiplied then rounded to a countable whole or half unit (e.g. eggs, spoons)unscaled— left as-is (e.g. pinches, salt, no-quantity lines)
This avoids nonsense results like "2.4 eggs" and never rounds down to zero.
Scale any ingredient list offline (
scale_ingredients): Apply the same scaling logic to any pasted French ingredient list (up to 100 lines) without a network request. Provide a directfactoror afrom_servings/to_servingspair.
Design & Safety
Fully read-only — never posts anything to Marmiton.
Robust parsing — uses structured
schema.orgJSON-LD rather than CSS scraping, so it's resilient to website redesigns.Polite access — honours
robots.txt, caches responses (15 min / 200 entries), and retries with backoff.Clear error handling —
rate_limitedandparse_failureerrors are surfaced with guidance, rather than silently returning nothing.Attribution-friendly — every result includes the recipe title, source URL, and an attribution string.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-marmitonFind a recipe for tarte aux pommes and scale it to 4 servings."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-marmiton
Marmiton is the largest French cooking site, where home cooks have been publishing their recipes since 1999. Each one gives its ingredients with their quantities, the steps to follow, the preparation and cooking times, the number of servings it is written for, and the ratings the people who made it left behind.
This server connects a chat client to that site. You can search its recipes by dish or by ingredient, read one in full with its ingredients and its steps, and rescale the quantities to the number of people at your table, with each line saying whether the figure is exact or was moved to stay usable in a kitchen. It needs no API key and no account.
Install
One-click install
Claude Code
claude mcp add marmiton -- npx -y mcp-marmitonClaude Desktop, Cursor, and any client using the standard config format
{
"mcpServers": {
"marmiton": {
"command": "npx",
"args": ["-y", "mcp-marmiton"]
}
}
}Node 24 or later is required, and no environment variable has to be set.
With Docker
{
"mcpServers": {
"marmiton": {
"command": "docker",
"args": ["run", "-i", "--rm", "ghcr.io/smeet666/mcp-marmiton:2.0.1"]
}
}
}-i keeps stdin open, which is where the protocol travels, and -t is left out
because a TTY rewrites the stream. The container needs outbound HTTPS to
www.marmiton.org, and nothing else: no volume, no port, no credential.
Bundle, without npm
Download mcp-marmiton-2.0.1.mcpb from
the latest release
and open it. A client that supports MCP bundles installs it on its own, with no
npm and no configuration file to edit. The bundle carries its dependencies, so
nothing is fetched at install time.
Related MCP server: cookwith-mcp
What you can ask
« Trouve-moi une recette de tarte aux pommes. »
"Read me that recipe for six people instead of four."
"What can I make with courgettes and chèvre?"
"Here is a recipe I copied from a book, scale it by 1.5 for me."
"How long does the second one take to cook?"
Marmiton is a French site, so its recipes are found in French: tarte aux pommes, poulet curry coco. The ordinary path runs from a search to a reading:
search_recipes names an id, and get_recipe takes that id.
Tools
Tool | What it does |
| Finds recipes by dish or by ingredient. |
| Reads one recipe, rescaled to a number of servings on request. |
| Rescales any ingredient list, with no request to the site. |
The server only reads. It publishes nothing to Marmiton.
search_recipes
Searches the recipes for a dish or an ingredient. Marmiton matches on the opening letters of a word, so a query brings back what the site ranked for it, and the answer says when the titles carry none of the words asked for.
Argument | Type | Required | What it does |
| string, up to 200 characters | yes | What to search for, in French. |
| integer, 1 to 30, default | no | Recipes to serve from this page of results. |
In return: rows carrying id, which get_recipe takes; title; url; and
image_url, which is null for a recipe published without a picture. Alongside
come result_count and total_available, the recipes on this page before
limit was applied. Marmiton's robots.txt disallows paging through search
results, so one page is what a search reads: narrow the query to see other
recipes.
get_recipe
Reads one recipe in full, and rescales its quantities when a number of servings is given.
Argument | Type | Required | What it does |
| string of digits | one of two | The Marmiton recipe id, as |
| a marmiton.org URL | one of two | The address of the recipe, used when |
| number, above 0 and up to 500 | no | Rescale the quantities to this many servings. |
In return: title, url, ingredients, steps, prep_minutes,
cook_minutes, total_minutes, category, author, rating and nutrition,
each null when the page publishes none. yield says what the recipe was
written for and what it was rescaled to: original_count, original_text,
requested, unit for what is being counted, and factor for the multiplier
applied. Every ingredient carries original, text, amount, amountMax for a
range, unit, and scaling, which reads scaled, rounded or unscaled. The
figures are this server's arithmetic, so say they were recomputed when you show
them. nutrition describes the recipe as published, at its own number of
servings.
scale_ingredients
Applies the same arithmetic to any list of ingredient lines, with no request to the site, so it works on a recipe copied from a book or a family notebook.
Argument | Type | Required | What it does |
| array of 1 to 100 strings, up to 300 chars | yes | The lines to rescale, in French. |
| number, above 0 and up to 100 | one of two | The multiplier to apply. |
| number, above 0 and up to 500 | one of two | How many servings the list is written for. |
| number, above 0 and up to 500 | one of two | How many servings are wanted. |
Pass factor, or the from_servings and to_servings pair.
In return: the factor used, the rescaled ingredients in the shape
get_recipe returns, and scaled_count, rounded_count and unscaled_count,
which count the lines whose value the rounding moved.
Scaling the quantities
Every ingredient comes back with a scaling flag saying what the rescaling could
do with its quantity.
Flag | Meaning | Example |
| The value is the product itself. |
|
| The value was moved to stay usable. |
|
| Carries no quantity, so left exactly as written. |
|
A quantity is stated in the unit that suits it, so a line can come back in a
different unit from the one the recipe used: 200 g multiplied by twenty reads
4 kg.
How finely an ingredient can be divided depends on what it is. A baguette can be
cut in two, in three or in four; an egg cannot be shared out. A quantity landing
between the two is rounded, and the rescaled recipe then departs a little from
the proportions of the original. The line carries rounded, and its note says
what was done.
Configuration
Every variable is optional. Set them in the env block of your client config.
Variable | Default | What it does |
| the project identity | Names your application to Marmiton, with an address where a person can be reached. |
|
| Gap between two requests, from 500 to 60000. A figure under the floor is refused and this one used. |
|
| Deadline for one request, from 1000 to 120000. |
|
| Attempts after a transient failure, from 0 to 10. |
|
| How long a page stays in memory, from 0 to 86400000. |
|
| Pages held in memory at once, from 0 to 10000. |
|
|
|
A value outside its range falls back to the default, and the reason is written to stderr.
Errors
Every failure carries one of six codes, a message, and where it helps a hint naming the next move.
Code | What happened | What to do |
| Marmiton answered, and holds no such recipe. | Check the id with |
| The arguments were refused before any request went out. | Read the message, which names the argument. |
| Marmiton asked this client to slow down. | Wait the number of seconds the hint names and call again with the same arguments. The recipe is still there. |
| The page loaded and the expected content was absent. | Report it at the issue tracker. |
| The request did not complete. | Try again shortly. |
| The request passed its deadline. | Raise |
As a library
The layer reading Marmiton is published on its own, with its pacing, its cache and its errors, and with no protocol attached.
import { MarmitonClient } from "mcp-marmiton/client";
const client = new MarmitonClient();
const { data, cached } = await client.getRecipe({ id: "18257" });
console.log(data.title, data.ingredients.length, cached);search and getRecipe each answer { data, cached }, and throw an error
carrying one of the six codes. The floor between two requests holds here as well.
Pacing and attribution
Requests go out one at a time with a minimum gap between them, and that floor
holds however the server is configured. The User-Agent always ends with the
project identity and an address where a person can be reached. Everything is read
from the schema.org JSON-LD Marmiton publishes for machines, and the paths its
robots.txt disallows are left alone.
Every result carries the title and the address of the recipe, and get_recipe
carries the author when the page names one, along with attribution, the title
and the address written as one line.
The recipes belong to Marmiton and to the cooks who wrote them. This MCP server is an unofficial project, with no affiliation to Marmiton.
Privacy
This server collects nothing about you and sends nothing to its author. It runs
on your machine, contacts www.marmiton.org and nothing else, holds its answers in memory
while it runs, and writes nothing to disk.
PRIVACY.md states what a request carries and which settings change
any of it.
Development
npm install
npm run build:fixtures
npm test
npm run checkTests run against generated fixtures and make no network request. The live suite,
npm run test:live, makes one request per route and runs nightly against the
site itself.
Contributing
Bugs, questions and ideas belong in the issue tracker. Pull requests are welcome; opening an issue first helps agree on the shape of the change. See CONTRIBUTING.md.
License
MIT, see LICENSE. The recipes belong to Marmiton and to their authors.
mcp-marmiton (français)
Marmiton est le plus grand site de cuisine français, où des cuisiniers publient leurs recettes depuis 1999. Chacune donne ses ingrédients avec leurs quantités, les étapes à suivre, les temps de préparation et de cuisson, le nombre de parts pour lequel elle est écrite, et les notes laissées par ceux qui l'ont faite.
Ce serveur relie un client de conversation à ce site. On peut y chercher des recettes par plat ou par ingrédient, en lire une en entier avec ses ingrédients et ses étapes, et adapter les quantités au nombre de convives, chaque ligne disant si le chiffre est exact ou s'il a été déplacé pour rester utilisable en cuisine. Aucune clé d'API, aucun compte.
Installation
Installation en un clic
Claude Code
claude mcp add marmiton -- npx -y mcp-marmitonClaude Desktop, Cursor, et tout client au format de configuration standard
{
"mcpServers": {
"marmiton": {
"command": "npx",
"args": ["-y", "mcp-marmiton"]
}
}
}Node 24 ou plus récent est nécessaire, et aucune variable d'environnement n'est à renseigner.
Avec Docker
{
"mcpServers": {
"marmiton": {
"command": "docker",
"args": ["run", "-i", "--rm", "ghcr.io/smeet666/mcp-marmiton:2.0.1"]
}
}
}-i garde l'entrée standard ouverte, qui est le canal du protocole, et -t est
omis parce qu'un TTY réécrit le flux. Le conteneur a besoin d'un accès HTTPS
sortant vers www.marmiton.org, et de rien d'autre : aucun volume, aucun port,
aucun identifiant.
Bundle, sans npm
Téléchargez mcp-marmiton-2.0.1.mcpb depuis
la dernière publication
et ouvrez-le. Un client qui gère les bundles MCP l'installe seul, sans npm et
sans fichier de configuration à modifier. Le bundle emporte ses dépendances, donc
rien n'est téléchargé à l'installation.
Ce qu'on peut demander
« Trouve-moi une recette de tarte aux pommes. »
« Lis-moi cette recette pour six personnes au lieu de quatre. »
« Qu'est-ce que je peux faire avec des courgettes et du chèvre ? »
« Voici une recette recopiée d'un livre, multiplie-la par 1,5. »
« Combien de temps de cuisson pour la deuxième ? »
Marmiton est un site français, donc ses recettes se trouvent en français : tarte aux pommes, poulet curry coco. Le chemin ordinaire va d'une recherche à une
lecture : search_recipes nomme un id, et get_recipe reprend cet
identifiant.
Les outils
Outil | Ce qu'il fait |
| Trouve des recettes par plat ou par ingrédient. |
| Lit une recette, adaptée à un nombre de parts sur demande. |
| Adapte n'importe quelle liste d'ingrédients, sans requête au site. |
Le serveur ne fait que lire. Il ne publie rien sur Marmiton.
search_recipes
Cherche des recettes par plat ou par ingrédient. Marmiton fait correspondre les premières lettres d'un mot, donc une requête ramène ce que le site a classé pour elle, et la réponse signale quand les titres ne portent aucun des mots demandés.
Argument | Type | Requis | Ce qu'il fait |
| chaîne, jusqu'à 200 caractères | oui | Ce qu'on cherche, en français. |
| entier, 1 à 30, défaut | non | Recettes à servir depuis cette page de résultats. |
En retour : des lignes portant id, que get_recipe reprend ; title ;
url ; et image_url, null pour une recette publiée sans photo. Viennent
aussi result_count et total_available, les recettes de cette page avant
l'application de limit. Le robots.txt de Marmiton interdit de paginer les
résultats de recherche, donc une recherche lit une page : resserrez la requête
pour voir d'autres recettes.
get_recipe
Lit une recette en entier, et adapte ses quantités quand un nombre de parts est donné.
Argument | Type | Requis | Ce qu'il fait |
| chaîne de chiffres | l'un des deux | L'identifiant Marmiton rendu par |
| une adresse marmiton.org | l'un des deux | L'adresse de la recette, utilisée à défaut d' |
| nombre, au-delà de 0 jusqu'à 500 | non | Adapte les quantités à ce nombre de parts. |
En retour : title, url, ingredients, steps, prep_minutes,
cook_minutes, total_minutes, category, author, rating et nutrition,
chacun null quand la page n'en publie pas. yield dit pour quoi la recette est
écrite et vers quoi elle a été adaptée : original_count, original_text,
requested, unit pour ce qui est compté, et factor pour le multiplicateur
appliqué. Chaque ingrédient porte original, text, amount, amountMax pour
un intervalle, unit, et scaling, qui vaut scaled, rounded ou unscaled.
Les chiffres sont l'arithmétique de ce serveur, donc dites qu'ils ont été
recalculés quand vous les montrez. nutrition décrit la recette telle que
publiée, pour son propre nombre de parts.
scale_ingredients
Applique la même arithmétique à n'importe quelle liste d'ingrédients, sans requête au site, donc sur une recette recopiée d'un livre ou d'un carnet de famille.
Argument | Type | Requis | Ce qu'il fait |
| tableau de 1 à 100 chaînes, jusqu'à 300 caractères | oui | Les lignes à adapter, en français. |
| nombre, au-delà de 0 jusqu'à 100 | l'un des deux | Le multiplicateur à appliquer. |
| nombre, au-delà de 0 jusqu'à 500 | l'un des deux | Le nombre de parts pour lequel la liste est écrite. |
| nombre, au-delà de 0 jusqu'à 500 | l'un des deux | Le nombre de parts voulu. |
Passez factor, ou le couple from_servings et to_servings.
En retour : le factor employé, les ingredients adaptés dans la forme que
rend get_recipe, et scaled_count, rounded_count et unscaled_count, qui
comptent les lignes dont l'arrondi a déplacé la valeur.
L'adaptation des quantités
Chaque ingrédient revient avec un drapeau scaling qui dit ce que l'adaptation a
pu faire de sa quantité.
Drapeau | Ce que cela veut dire | Exemple |
| La valeur est le produit lui-même. |
|
| La valeur a été déplacée pour rester utilisable. |
|
| Ne porte aucune quantité, laissée telle quelle. |
|
Une quantité est exprimée dans l'unité qui lui convient. Après adaptation, une
ligne peut donc apparaître dans une autre unité que celle de la recette : 200 g
multipliés par vingt donnent 4 kg.
La finesse à laquelle un ingrédient se coupe dépend de sa nature. Une baguette se
coupe en deux, en trois ou en quatre ; un oeuf ne se partage pas. Une quantité
qui tombe entre les deux est donc arrondie, et la recette adaptée s'écarte alors
un peu des proportions de l'originale. La ligne porte rounded, et sa note dit
ce qui a été fait.
Configuration
Chaque variable est facultative. Elles se posent dans le bloc env de la
configuration du client.
Variable | Défaut | Ce qu'elle fait |
| l'identité du projet | Nomme votre application auprès de Marmiton, avec une adresse où joindre une personne. |
|
| Écart entre deux requêtes, de 500 à 60000. Une valeur sous le plancher est refusée au profit de celle-ci. |
|
| Délai d'une requête, de 1000 à 120000. |
|
| Tentatives après un échec passager, de 0 à 10. |
|
| Durée pendant laquelle une page reste en mémoire, de 0 à 86400000. |
|
| Pages gardées en mémoire à la fois, de 0 à 10000. |
|
|
|
Une valeur hors de sa plage retombe sur le défaut, et la raison est écrite sur la sortie d'erreur.
Erreurs
Chaque échec porte un des six codes, un message, et quand cela aide une indication du geste suivant.
Code | Ce qui s'est passé | Que faire |
| Marmiton a répondu, et n'a pas cette recette. | Vérifiez l'identifiant avec |
| Les arguments ont été refusés avant toute requête. | Lisez le message, qui nomme l'argument. |
| Marmiton demande à ce client de ralentir. | Attendez les secondes indiquées et rappelez avec les mêmes arguments. La recette est toujours là. |
| La page a chargé et le contenu attendu est absent. | Signalez-le sur le suivi d'incidents. |
| La requête n'a pas abouti. | Réessayez sous peu. |
| La requête a dépassé son délai. | Augmentez |
Comme bibliothèque
La couche qui lit Marmiton est publiée seule, avec son rythme, son cache et ses erreurs, sans protocole attaché.
import { MarmitonClient } from "mcp-marmiton/client";
const client = new MarmitonClient();
const { data, cached } = await client.getRecipe({ id: "18257" });
console.log(data.title, data.ingredients.length, cached);search et getRecipe répondent chacun { data, cached }, et lèvent une erreur
portant un des six codes. Le plancher entre deux requêtes tient également ici.
Rythme et attribution
Les requêtes partent une à une avec un écart minimal entre elles, et ce plancher
tient quelle que soit la configuration. Le User-Agent se termine toujours par
l'identité du projet et une adresse où joindre une personne. Tout est lu dans le
JSON-LD schema.org que Marmiton publie pour les machines, et les chemins que
son robots.txt interdit sont laissés tranquilles.
Chaque résultat porte le titre et l'adresse de la recette, et get_recipe porte
l'auteur quand la page le nomme, ainsi qu'attribution, le titre et l'adresse
écrits en une ligne.
Les recettes appartiennent à Marmiton et aux cuisiniers qui les ont écrites. Ce MCP est un projet non officiel, sans affiliation à Marmiton.
Confidentialité
Ce serveur ne collecte rien sur vous et n'envoie rien à son auteur. Il tourne sur
votre machine, ne joint que www.marmiton.org, garde ses réponses en mémoire le temps qu'il
tourne, et n'écrit rien sur le disque. PRIVACY.md dit ce qu'une
requête emporte et quels réglages changent cela.
Développement
npm install
npm run build:fixtures
npm test
npm run checkLes tests s'exécutent sur des fixtures engendrées et n'émettent aucune requête.
La suite en direct, npm run test:live, émet une requête par route et tourne
chaque nuit contre le site lui-même.
Contribuer
Les anomalies, les questions et les idées ont leur place dans le suivi d'incidents. Les propositions de modification sont bienvenues ; ouvrir un ticket d'abord aide à s'accorder sur la forme du changement. Voir CONTRIBUTING.md.
Licence
MIT, voir LICENSE. Les recettes appartiennent à Marmiton et à leurs auteurs.
Available Tools
3 toolsget_recipeGet a recipeARead-onlyIdempotent
Read one Marmiton recipe: ingredients, steps, times, category, rating and nutrition. Give the id returned by search_recipes, or a marmiton.org recipe URL. Set 'servings' to rescale the quantities. Each ingredient reports how it was handled: 'scaled' when the arithmetic came out exact, which a count of eggs, spoons or pinches reaches as readily as a mass in grams, 'rounded' when the value had to be moved to stay usable, and 'unscaled' for lines carrying no quantity at all. Trust that flag rather than recomputing: the point of scaling here is to avoid answers like '2.4 eggs'. Recipes yielding pieces rather than servings are rescaled the same way; check 'yield' to see which. Always cite 'attribution' when showing a recipe to a user.
| Name | Required | Description | Default |
|---|---|---|---|
| id | No | Marmiton recipe id, as returned by search_recipes. Preferred over 'url'. | |
| url | No | Full marmiton.org recipe URL. Only marmiton.org is accepted. Ignored when 'id' is given. | |
| servings | No | Rescale the ingredients to this many servings. Omit to get the recipe exactly as published. |
Output Schema
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| url | Yes | |
| notes | Yes | |
| steps | Yes | |
| title | Yes | |
| yield | Yes | |
| author | Yes | |
| rating | Yes | |
| source | Yes | |
| category | Yes | |
| nutrition | Yes | Nutrition is given for the recipe as published, not rescaled. |
| attribution | Yes | |
| ingredients | Yes | |
| cook_minutes | Yes | |
| prep_minutes | Yes | |
| total_minutes | Yes |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
While annotations already indicate read-only, idempotent, and non-destructive, the description adds important behavioral details such as how scaling flags work ('scaled', 'rounded', 'unscaled'), the rationale to avoid non-integer quantities, and the yield field for pieces. This goes beyond the annotations, though it doesn't cover all edge cases like error handling or rate limits, but the additions are substantial.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is fairly long but each sentence provides necessary detail, especially the scaling explanation, which is critical for correct usage. It front-loads the core purpose and then dives into specifics. The length is justified, though it could be slightly more streamlined, but it's effective and not wasteful.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity, the description covers the key aspects: input parameters, scaling behavior, yield, and citation requirement. The output schema likely details return structure, so the description doesn't need to repeat that. Annotations cover safety, and the description fills in usage nuances. A few edge cases (e.g., what happens if id not found) aren't mentioned, but overall it's comprehensive for the use case.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3, but the description adds extra semantics for the 'servings' parameter, explaining the scaling behavior and the flag system. For 'id' and 'url', it clarifies the preference and precedence rules beyond what the schema states (e.g., 'Preferred over url', 'Ignored when id is given'). This adds meaningful value, justifying a 4.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it reads a single Marmiton recipe with specific content (ingredients, steps, times, category, rating, nutrition). It distinguishes itself from sibling tools by referencing search_recipes for obtaining the id and by handling scaling differently from the separate scale_ingredients tool.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicit guidance on when to use this tool: provides the id from search_recipes or a marmiton.org URL, with clear alternatives and exclusions (URL ignored if id given, only marmiton.org accepted). It also instructs when to omit servings and when to use scaling flags, and mentions citing attribution, which effectively covers usage context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
scale_ingredientsScale an ingredient listARead-onlyIdempotent
Rescale a list of ingredient lines to a different number of servings, without contacting any website. Give either 'factor' directly, or 'from_servings' and 'to_servings' and the factor is computed. Works on any French ingredient list, whatever its source, so it also serves a recipe the user pasted in. Quantities in grams or millilitres are multiplied and rounded to readable values; a countable thing lands on a half when half of one can be poured, weighed or cut, as a boîte, a sachet, a feuille de gélatine or a cuillère can, and on a whole one when it cannot, as an oeuf, a jaune or a blanc; approximate measures such as a pinch or a handful have their count multiplied in whole units and stay in their own vocabulary. A line writing an article where a digit would go, as in 'un bouchon de rhum' or 'une pincée de sel', is read as one of that measure. Lines carrying no quantity are returned untouched and flagged. Prefer this over doing the arithmetic yourself.
| Name | Required | Description | Default |
|---|---|---|---|
| factor | No | Multiplier to apply. Use this or the from/to pair. | |
| ingredients | Yes | Ingredient lines, for example ['200 g de farine', '3 oeufs', 'sel']. | |
| to_servings | No | How many servings are wanted. | |
| from_servings | No | How many servings the list is written for. |
Output Schema
| Name | Required | Description |
|---|---|---|
| notes | Yes | |
| factor | Yes | |
| ingredients | Yes | |
| scaled_count | Yes | |
| rounded_count | Yes | Lines whose value rounding moved away from the exact product, not lines that could have been rounded. |
| unscaled_count | Yes |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description discloses detailed behavior: quantities are multiplied and rounded, countable items are handled with fractions, approximate measures stay whole, and lines without quantities are flagged. This goes beyond annotations (readOnly, idempotent) and fully informs the user of the tool's effects.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is overly verbose and repetitive. For example, the sentence about countable items and approximate measures is long, and the final sentence about 'A line writing an article where a digit would go' restates earlier content. This could be condensed without losing meaning.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers all necessary aspects: what the tool does, how to specify scaling, handling of different unit types, and behavior for missing quantities. It is complete enough for an agent to use the tool effectively without additional clarification.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description adds meaning to parameters: it explains the relationship between factor and from_servings/to_servings, and clarifies that ingredients are lines that may contain quantities. It also describes how lines without quantities are treated, providing more context than the schema descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Rescale a list of ingredient lines to a different number of servings'. It also specifies the input type (French ingredient list) and distinguishes it from sibling tools (search_recipes, get_recipe) by focusing on scaling.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit usage instructions: 'Give either factor directly, or from_servings and to_servings'. It also gives examples and notes about rounding and edge cases, but does not explicitly state when to prefer this tool over siblings, though the purpose is clear enough.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
search_recipesSearch recipesARead-onlyIdempotent
Search Marmiton for recipes by free text: a dish, an ingredient, or both, for example 'tarte aux pommes' or 'poulet curry coco'. Marmiton is a French site, so French terms work best. Returns candidate recipes with the id and URL needed to read one with get_recipe. One call returns one page of results. Marmiton's robots.txt disallows paginating search results, so this server does not, and there is no page parameter: narrow the query instead of asking for more pages.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Maximum recipes to return from this page of results. | |
| query | Yes | What to search for, in French, for example 'tarte aux pommes'. |
Output Schema
| Name | Required | Description |
|---|---|---|
| notes | Yes | |
| query | Yes | |
| source | Yes | |
| results | Yes | |
| result_count | Yes | |
| total_available | Yes | Recipes on this page before applying 'limit'. |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, openWorldHint, idempotentHint, and destructiveHint false, and the description does not contradict them. It adds valuable context about the one-page result and robots.txt restriction, plus the return of ids and URLs, enriching the behavioral picture beyond the annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and well-structured, with each sentence serving a distinct purpose: purpose, examples, result guidance, and pagination constraint. There is no redundancy or fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's simplicity and the presence of an output schema, the description covers all necessary aspects: search intent, language considerations, result format, and pagination constraints. It is complete for an agent to understand and use the tool effectively.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema fully describes both parameters, but the description enriches the query parameter with examples and language guidance (e.g., French terms work best). It also clarifies the limit's role by explaining the lack of pagination, adding meaning beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool searches recipes on Marmiton by free text with a specific verb and resource. It distinguishes from sibling tools by mentioning it returns IDs and URLs used by get_recipe, making its role as the search entry point explicit.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context that Marmiton is a French site and advises using French terms. It explicitly explains the single-page limitation and advises narrowing queries instead of expecting pagination, effectively covering when and how to use the tool, though it does not explicitly state when not to use it.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
3 tool updates
- Changed
get_recipe23 fields changed- added
Output schema / properties / author / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / author / typeRemoved value: -[ - "string", - "null" -] - added
Output schema / properties / category / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / category / typeRemoved value: -[ - "string", - "null" -] - changed
Output schema / properties / cook_minutes / anyOfPrevious value: -[ - { - "type": "integer" - }, - { - "type": "null" - } -]New value: +[ + { + "maximum": 9007199254740991, + "minimum": -9007199254740991, + "type": "integer" + }, + { + "type": "null" + } +] - added
Output schema / properties / ingredients / items / properties / amount / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / amount / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / ingredients / items / properties / amountMax / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / amountMax / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / ingredients / items / properties / unit / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / unit / typeRemoved value: -[ - "string", - "null" -] - changed
Output schema / properties / nutrition / anyOfPrevious value: -[ - { - "additionalProperties": false, - "properties": { - "calories": { - "type": [ - "string", - "null" - ] - }, - "carbohydrate": { - "type": [ - "string", - "null" - ] - }, - "fat": { - "type": [ - "string", - "null" - ] - }, - "fiber": { - "type": [ - "string", - "null" - ] - }, - "protein": { - "type": [ - "string", - "null" - ] - }, - "serving_size": { - "type": [ - "string", - "null" - ] - }, - "sodium": { - "type": [ - "string", - "null" - ] - } - }, - "required": [ - "calories", - "protein", - "fat", - "carbohydrate", - "fiber", - "sodium", - "serving_size" - ], - "type": "object" - }, - { - "type": "null" - } -]New value: +[ + { + "additionalProperties": false, + "properties": { + "calories": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "carbohydrate": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "fat": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "fiber": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "protein": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "serving_size": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + }, + "sodium": { + "anyOf": [ + { + "type": "string" + }, + { + "type": "null" + } + ] + } + }, + "required": [ + "calories", + "protein", + "fat", + "carbohydrate", + "fiber", + "sodium", + "serving_size" + ], + "type": "object" + }, + { + "type": "null" + } +] - changed
Output schema / properties / prep_minutes / anyOfPrevious value: -[ - { - "type": "integer" - }, - { - "type": "null" - } -]New value: +[ + { + "maximum": 9007199254740991, + "minimum": -9007199254740991, + "type": "integer" + }, + { + "type": "null" + } +] - changed
Output schema / properties / rating / anyOfPrevious value: -[ - { - "additionalProperties": false, - "properties": { - "best": { - "type": [ - "number", - "null" - ] - }, - "count": { - "type": [ - "number", - "null" - ] - }, - "value": { - "type": "number" - } - }, - "required": [ - "value", - "count", - "best" - ], - "type": "object" - }, - { - "type": "null" - } -]New value: +[ + { + "additionalProperties": false, + "properties": { + "best": { + "anyOf": [ + { + "type": "number" + }, + { + "type": "null" + } + ] + }, + "count": { + "anyOf": [ + { + "type": "number" + }, + { + "type": "null" + } + ] + }, + "value": { + "type": "number" + } + }, + "required": [ + "value", + "count", + "best" + ], + "type": "object" + }, + { + "type": "null" + } +] - changed
Output schema / properties / total_minutes / anyOfPrevious value: -[ - { - "type": "integer" - }, - { - "type": "null" - } -]New value: +[ + { + "maximum": 9007199254740991, + "minimum": -9007199254740991, + "type": "integer" + }, + { + "type": "null" + } +] - added
Output schema / properties / yield / properties / factor / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / yield / properties / factor / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / yield / properties / original_count / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / yield / properties / original_count / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / yield / properties / requested / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / yield / properties / requested / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / yield / properties / unit / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / yield / properties / unit / typeRemoved value: -[ - "string", - "null" -]
- Changed
scale_ingredients12 fields changed- added
Output schema / properties / ingredients / items / properties / amount / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / amount / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / ingredients / items / properties / amountMax / anyOfAdded value: +[ + { + "type": "number" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / amountMax / typeRemoved value: -[ - "number", - "null" -] - added
Output schema / properties / ingredients / items / properties / unit / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / ingredients / items / properties / unit / typeRemoved value: -[ - "string", - "null" -] - added
Output schema / properties / rounded_count / maximumAdded value: +9007199254740991 - added
Output schema / properties / rounded_count / minimumAdded value: +-9007199254740991 - added
Output schema / properties / scaled_count / maximumAdded value: +9007199254740991 - added
Output schema / properties / scaled_count / minimumAdded value: +-9007199254740991 - added
Output schema / properties / unscaled_count / maximumAdded value: +9007199254740991 - added
Output schema / properties / unscaled_count / minimumAdded value: +-9007199254740991
- Changed
search_recipes6 fields changed- added
Output schema / properties / result_count / maximumAdded value: +9007199254740991 - added
Output schema / properties / result_count / minimumAdded value: +-9007199254740991 - added
Output schema / properties / results / items / properties / image_url / anyOfAdded value: +[ + { + "type": "string" + }, + { + "type": "null" + } +] - removed
Output schema / properties / results / items / properties / image_url / typeRemoved value: -[ - "string", - "null" -] - added
Output schema / properties / total_available / maximumAdded value: +9007199254740991 - added
Output schema / properties / total_available / minimumAdded value: +-9007199254740991
2 tool updates
v1.7.0- Changed
get_recipe3 fields changed- added
Output schema / properties / ingredients / items / properties / amountMaxAdded value: +{ + "description": "Upper bound when the line gives a range, as in '200 à 300 g', with 'amount' holding the lower one. Null when the line states a single amount.", + "type": [ + "number", + "null" + ] +} - changed
Output schema / properties / ingredients / items / properties / scaling / descriptionPrevious value: -"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pinch multiplied by six is six pinches. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left alone, because the line carries no quantity."New value: +"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pincée multiplied by six is six pincées. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left as published, because nothing on the line is the factor's to multiply: the note says which of the two it is." - changed
Output schema / properties / ingredients / items / requiredPrevious value: -[ - "original", - "text", - "amount", - "unit", - "scaling", - "adjusted" -]New value: +[ + "original", + "text", + "amount", + "amountMax", + "unit", + "scaling", + "adjusted" +]
- Changed
scale_ingredients3 fields changed- added
Output schema / properties / ingredients / items / properties / amountMaxAdded value: +{ + "description": "Upper bound when the line gives a range, as in '200 à 300 g', with 'amount' holding the lower one. Null when the line states a single amount.", + "type": [ + "number", + "null" + ] +} - changed
Output schema / properties / ingredients / items / properties / scaling / descriptionPrevious value: -"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pinch multiplied by six is six pinches. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left alone, because the line carries no quantity."New value: +"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pincée multiplied by six is six pincées. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left as published, because nothing on the line is the factor's to multiply: the note says which of the two it is." - changed
Output schema / properties / ingredients / items / requiredPrevious value: -[ - "original", - "text", - "amount", - "unit", - "scaling", - "adjusted" -]New value: +[ + "original", + "text", + "amount", + "amountMax", + "unit", + "scaling", + "adjusted" +]
2 tool updates
v1.5.0- Changed
get_recipe3 fields changed- changed
Output schema / properties / ingredients / items / properties / adjusted / descriptionPrevious value: -"True when rounding moved the value away from the exact product. A line can be 'rounded' and still land exactly, as three eggs doubled land on six."New value: +"True when rounding moved the value away from the exact product, which is what makes a line 'rounded' rather than 'scaled'." - changed
Output schema / properties / ingredients / items / properties / note / descriptionPrevious value: -"Why the line was rounded, clamped or left alone."New value: +"Why the line was rounded, clamped or left alone, including when the measure is approximate and only its count was multiplied." - changed
Output schema / properties / ingredients / items / properties / scaling / descriptionPrevious value: -"'scaled' was multiplied cleanly. 'rounded' was multiplied then rounded to something countable. 'unscaled' was left alone, either because it carries no quantity or because the measure is approximate by nature."New value: +"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pinch multiplied by six is six pinches. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left alone, because the line carries no quantity."
- Changed
scale_ingredients3 fields changed- changed
Output schema / properties / ingredients / items / properties / adjusted / descriptionPrevious value: -"True when rounding moved the value away from the exact product. A line can be 'rounded' and still land exactly, as three eggs doubled land on six."New value: +"True when rounding moved the value away from the exact product, which is what makes a line 'rounded' rather than 'scaled'." - changed
Output schema / properties / ingredients / items / properties / note / descriptionPrevious value: -"Why the line was rounded, clamped or left alone."New value: +"Why the line was rounded, clamped or left alone, including when the measure is approximate and only its count was multiplied." - changed
Output schema / properties / ingredients / items / properties / scaling / descriptionPrevious value: -"'scaled' was multiplied cleanly. 'rounded' was multiplied then rounded to something countable. 'unscaled' was left alone, either because it carries no quantity or because the measure is approximate by nature."New value: +"'scaled' means the arithmetic came out exact, which a count reaches as readily as a mass: one pinch multiplied by six is six pinches. 'rounded' means the value had to be moved to stay usable, to a whole or half unit or to a step a scale can show. 'unscaled' was left alone, because the line carries no quantity."
2 tool updates
v1.3.0- Changed
get_recipe3 fields changed- added
Output schema / properties / ingredients / items / properties / adjustedAdded value: +{ + "description": "True when rounding moved the value away from the exact product. A line can be 'rounded' and still land exactly, as three eggs doubled land on six.", + "type": "boolean" +} - changed
Output schema / properties / ingredients / items / properties / note / descriptionPrevious value: -"Why the line was rounded or left alone."New value: +"Why the line was rounded, clamped or left alone." - changed
Output schema / properties / ingredients / items / requiredPrevious value: -[ - "original", - "text", - "amount", - "unit", - "scaling" -]New value: +[ + "original", + "text", + "amount", + "unit", + "scaling", + "adjusted" +]
- Changed
scale_ingredients4 fields changed- added
Output schema / properties / ingredients / items / properties / adjustedAdded value: +{ + "description": "True when rounding moved the value away from the exact product. A line can be 'rounded' and still land exactly, as three eggs doubled land on six.", + "type": "boolean" +} - changed
Output schema / properties / ingredients / items / properties / note / descriptionPrevious value: -"Why the line was rounded or left alone."New value: +"Why the line was rounded, clamped or left alone." - changed
Output schema / properties / ingredients / items / requiredPrevious value: -[ - "original", - "text", - "amount", - "unit", - "scaling" -]New value: +[ + "original", + "text", + "amount", + "unit", + "scaling", + "adjusted" +] - added
Output schema / properties / rounded_count / descriptionAdded value: +"Lines whose value rounding moved away from the exact product, not lines that could have been rounded."
3 tool updates
v1.0.2- First observed
get_recipe - First observed
scale_ingredients - First observed
search_recipes
TDQS
search_recipes and get_recipe are clearly separated: one searches, the other reads a specific recipe. A slight overlap exists because get_recipe can also rescale quantities via its servings parameter while scale_ingredients exists solely for rescaling, but the offline and user-supplied use case for scale_ingredients keeps the boundary mostly clear.
All three tool names follow a consistent verb_noun snake_case pattern: search_recipes, get_recipe, scale_ingredients. There is no mixing of naming styles or vague verbs.
Three tools is a well-scoped count for this server's purpose: search for recipes, fetch a recipe's details, and rescale ingredient lists. Each tool has a distinct role in the intended workflow and none feel ornamental or duplicated.
The core search-to-fetch-to-scaling workflow is covered with no broken path between tools. A minor gap is that search results cannot be paginated and there is no category/filter-based browsing, though the description explicitly suggests narrowing queries as a workaround instead.
Maintenance
Related MCP Connectors
Unlock the power of food transparency with our Open Food Facts MCP server. Easily look up any food
MCP server for Meitre restaurant reservations.
Recipes MCP — wraps TheMealDB API (free tier, no auth)
- mcpOAuthcom.zomato
An MCP server that exposes functionalities to use Zomato's services.
Related MCP Servers
- AlicenseAqualityDmaintenanceAn MCP server that enables users to search for recipes and cooking instructions via the API Ninjas Recipe API. It supports querying specific dishes and provides paginated results for recipe discovery.1MIT
- AlicenseBqualityDmaintenanceAn MCP server that enables AI-powered recipe generation and transformation using natural language, supporting dietary restrictions, allergies, and nutritional goals.222MIT
- FlicenseAqualityDmaintenanceMCP server that provides cooking recipe lookup, classification, and weekly meal planning based on dietary restrictions and allergies.5-
- FlicenseNot gradedqualityBmaintenanceMCP server for discovering and retrieving meal recipes from TheMealDB API, enabling search, random meals, category/ingredient filtering, and full recipe details.-
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/smeet666/mcp-marmiton'
If you have feedback or need assistance with the MCP directory API, please join our Discord server