ForkMate

Log what you ate by talking to your AI assistant — calories and macros, completely free.

CommunityOnlineNo sign-inGlobalFreeRead-only

What it can do

    What data it sees

    Do you need an account

    No: the server works without sign-in

    Log what you ate by talking to your AI assistant — calories and macros, completely free.

    How to connect

    How to connect ForkMate to Claude

    The server address to paste into Claude settings:

    https://mcp.forkmate.ai/

    1. Open Claude (claude.ai or the desktop app).
    2. Go to Settings → Connectors.
    3. Click Add custom connector.
    4. Paste the server address copied below into Remote MCP server URL and click Add.
    5. In a chat, click + → Connectors and switch the new connector on.

    Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.

    Authorization

    After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.

    How to connect ForkMate to ChatGPT

    The server address to paste into ChatGPT settings:

    https://mcp.forkmate.ai/

    1. Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
    2. Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
    3. Open Settings → Connectors and click Create.
    4. Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
    5. Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
    6. In a new chat open + → Apps/Connectors and enable the connector.

    OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.

    Authorization

    On first use ChatGPT opens the service's sign-in window. Sign in and approve access.

    How to connect ForkMate to Cursor

    The server address to paste into Cursor settings:

    https://mcp.forkmate.ai/

    Fastest: click Open in Cursor below and confirm the prompt.

    Manually:

    1. In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
    2. Paste the JSON copied below into ~/.cursor/mcp.json (per project: .cursor/mcp.json in the repo root).
    3. Save the file. The server appears in the MCP list; authorize it there if asked.

    Authorization

    If the service needs sign-in, an authorize button appears next to the server in the MCP list.

    Server tool list (12)

    Raw names from tools/list. Only developers need these.

    whoamiDiagnostic: returns the authenticated user id and scopes.
    get_dayRead the user's food diary for a day (entries + calorie/macro totals). SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    get_rangeRead the user's diary across a date range, with per-day calorie/macro totals. SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    get_preferencesRead the user's saved dietary preferences so you can tailor logging and suggestions WITHOUT re-asking every chat: their diet style, a structured list of allergies to avoid (the big-9 major allergens), foods they dislike, and a typical-portion note. IMPORTANT: the allergen list is self-reported and is NOT a safety guarantee — always tell the user to check ingredient labels themselves; cross-contamination and gaps in food data are not captured (see the returned allergy_disclaimer). The `allergies` field covers the major US allergens ONLY; a user may have an allergen outside it (e.g. mustard, celery, corn, mollusks, barley/rye) — ask about those directly. NEVER treat the `dislikes` list as an allergy: it is a taste preference to de-prioritize, never a safety exclusion.
    log_mealLog what the user ate to their food diary. Parse the user's free text into items and, when you can, include estimated macros per item for accuracy. SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    update_mealCorrect a food already logged to the user's diary — fix a wrong calorie/macro value, quantity, or name, or move an entry to a different meal. Identify the entry by its `id` and `local_date` (both from get_day) and the food by its `item_index` within that entry's items[]. Only the fields you send change; the macros you send are MERGED onto the existing ones (so sending just `kcal` leaves protein/carb/fat as they were). This overwrites the value IN PLACE — there is no history of the previous value. Editing never moves an entry to another day (to do that, delete and re-log). SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    delete_mealDelete a food from the user's diary — remove one food from an entry (by `item_index`), or the whole entry (omit `item_index`). Identify the entry by its `id` and `local_date` (both from get_day). This is a TRUE removal: the data is gone, with NO server-side tombstone and no undo. Deleting the last food in an entry removes the entry. Safe to retry — deleting something already gone is a no-op success. SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    get_pantryRead the user's PANTRY — the foods they keep ON HAND (their staples), so you can suggest meals from what they actually have and pre-fill macros when they log one. Returns each item's name and, when the user saved them, macros (for the item's serving), a serving label, a `source`, and a short note. The pantry is the user's CURATED list of what they stock — separate from what they've logged (their diary) and from their frequents (what they log often). IMPORTANT: a `source` (e.g. 'usda') is the user's own CLAIM about where the macros came from, NOT a server-verified guarantee — treat it as a hint, never as certified.
    add_pantry_itemAdd a food to the user's pantry, or UPDATE it if it's already there (matched by name, any casing) — e.g. 'add rolled oats to my pantry'. Only `name` is required; include `macros` (for one serving), a `serving` label, a `source`, and a short `note` when you know them, so a later log can reuse them. Re-adding the same food REPLACES its details (an upsert — it never creates a duplicate). Only pass a `source` you actually got from search_foods/lookup_barcode; an unrecognized value is recorded as the user's own estimate ('client'). This does NOT log a meal — it only curates the user's staples.
    remove_pantry_itemRemove a food from the user's pantry by name — e.g. 'take eggs off my pantry list'. This removes it from their on-hand STAPLES only; it does NOT delete anything from their food diary. Safe to retry — removing something that isn't in the pantry is a no-op success.
    search_foodsSearch USDA FoodData Central and Open Food Facts for foods matching a query, returning candidates with macros and a `source` you can show the user. IMPORTANT: the macros are PER 100 g (see each candidate's `serving`) — scale them to the portion the user actually ate before logging with log_meal. A candidate MAY also carry `serving_grams`/`serving_label` for ONE household serving (e.g. 48 g / "1 frank") — when present, offer the user 'N servings' instead of asking for grams, but still scale the per-100 g macros to the resolved grams before logging. When you log a chosen candidate, pass its `source` to log_meal so the diary records real provenance (USDA/Open Food Facts) instead of an estimate. SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.
    lookup_barcodeLook up a packaged food by its UPC/EAN barcode via Open Food Facts. IMPORTANT: the macros are PER 100 g (see `serving`) — scale to the portion eaten before logging with log_meal. It MAY also carry `serving_grams`/`serving_label` for one household serving — offer 'N servings' when present, still scaling the per-100 g macros before logging. Pass the returned `source` to log_meal to preserve provenance. SAFETY: all calorie and macro values here — including carbohydrates — are ESTIMATES (from USDA / Open Food Facts or the user's own entry), approximate and not lab-measured or per-batch. They are for general nutrition tracking ONLY. Do NOT use them for insulin dosing, carb-counting for a bolus, blood-glucose prediction, or any other medical decision, and do NOT compute a dose or glucose estimate from them. For anything medical, direct the user to the product's own label and their care team.