DataBody MCP — Claude Desktop integration

01WHAT IS MCP

An open standard. Real tool calls, not copy-paste.

The Model Context Protocol is an open standard that lets AI assistants like Claude connect to external tools and data sources. Instead of pasting between apps, you talk to DataBody directly through natural conversation.

modelcontextprotocol.io
open standard

Natural language

Tell Claude what you ate. It parses, estimates macros, and logs to your actual DataBody account.

▸ "had eggs + avocado toast" ↳ log_food · 542 kcal · 24P

Real actions

Not chat-only. Claude executes tools against the DataBody API — writes, deletes, updates, queries.

▸ record_checkin({mood: 8}) ↳ saved · 2026-04-24

Secure auth

OAuth 2.0 with PKCE. Token stored locally in ~/.databody_token.json. Revokable from the app.

▸ authenticate({method: "browser"}) ↳ authed · [email protected]

02INSTALL

Four steps. Three install paths.

The fastest route is the MCPB bundle — double-click to install. Power users can go via npm or clone the source. All three land in the same place: 60+ DataBody tools in your Claude sidebar.

claude_desktop_config.json

Prerequisites

Claude Desktop installed. Node.js 20+ (only if using npm/source). A DataBody account.

▸ node --version ↳ v20.9.0

Install server · MCPB

RECOMMENDED Download one file. Double-click. Done. No npm, no Node.js required on your system.

▸ databody-mcp.mcpb ↳ drag onto Claude Desktop

02b

Install server · npm

If you have Node.js, install globally. Updates are a single npm update -g.

▸ npm install -g @databody/mcp

02c

Install server · source

For developers customizing or contributing. Clone, install, build.

▸ git clone databody-ai/databody-mcp ▸ cd databody-mcp ▸ npm install && npm run build

Configure Claude Desktop

MCPB is auto-configured. For npm/source, edit claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\).

▸ open config → add mcpServers.databody ↳ see snippets below

Authenticate

Restart Claude Desktop. Ask it to connect. Browser OAuth opens. Token lands in ~/.databody_token.json.

▸ "Connect to DataBody" ↳ browser opens · ✓ authed

claude_desktop_config.json — npm install json

// Option B — if you installed via npm

{

"mcpServers": {

"databody": {

"command": "npx",

"args": ["-y", "@databody/mcp"],

"env": {

"DATABODY_API_URL": "https://databody.ai"

}

claude_desktop_config.json — source build json

// Option C — if you built from source

{

"mcpServers": {

"databody": {

"command": "node",

"args": ["/path/to/databody-mcp/dist/index.js"],

"env": {

"DATABODY_API_URL": "https://databody.ai"

}

~/ — authenticate claude

› Connect to DataBody

→ opening browser · oauth 2.0 + pkce

→ ✓ authed as [email protected]

→ token: ~/.databody_token.json

›

03EXAMPLES

Tell Claude what you want. It picks the tool.

No slash commands. No memorizing tool names. Claude reads your intent, selects from 60+ DataBody tools, and executes. You read the result in plain English.

natural language · any phrasing

Log a meal

Claude parses free text, estimates macros (including fiber, sugar, water), and logs it.

▸ "I had 2 scrambled eggs, avocado toast, and a coffee for breakfast" ↳ log_food · 542 kcal

Check progress

Summary of today vs targets: kcal, protein, carbs, fat, fiber, sugar, water.

▸ "How am I doing on my macros today?" ↳ get_today_nutrition

Get suggestions

AI meal suggestions based on remaining macros and fiber/sugar/water goals.

▸ "High-protein dinner with 600 kcal left?" ↳ get_meal_suggestions

Household planning

Reads member goals and suggests meals that work for everyone at the table.

▸ "Plan dinner for my family tonight" ↳ get_household_summary

Set preferences

Persistent notes that inform all future suggestions.

▸ "Remember that I'm lactose intolerant" ↳ create_note

Analyze photos

Vision identifies foods and estimates macros from an attached image.

▸ "What's in this meal?" [attach photo] ↳ analyze_photo

Grocery shopping

Create lists, add items, track spending, get smart suggestions per store.

▸ "Grocery list for Trader Joe's, $75 budget" ↳ create_grocery_list

Check off items

Send a photo of groceries and AI checks off matching items from your list.

▸ "I just picked these up" [photo] ↳ check_grocery_items

Daily check-in

Record 1–10 subjective scores: mood, energy, sleep, gut, hydration, skin, hangover.

▸ "Feeling 8/10 today, slept great" ↳ record_checkin

04TOOL REGISTRY

60+ tools. Organized by domain.

Claude picks the right one for each request. You don't need to remember names — but here they are. Every tool maps to a real endpoint on the DataBody API.

grouped by category
all documented

AUTHENTICATION

authenticatebrowser or password OAuth flow

›

logoutClear local token

›

auth_statusCheck current session

›

USER PROFILE

get_user_profileRead full profile

›

update_user_profileEdit fields

›

change_passwordRotate credentials

›

change_emailUpdate login address

›

HEALTH & SUMMARY

get_health_summaryDashboard snapshot

›

get_health_historyDate-range history

›

NUTRITION · kcal · protein · carbs · fat · fiber · sugar · water · caffeine · alcohol

get_today_nutritionToday's totals

›

get_nutrition_historyDate range logs

›

get_recent_intakeLast N entries

›

log_foodFrom text or voice

›

update_nutrition_logEdit a meal

›

delete_nutrition_logRemove a meal

›

add_nutrition_itemAppend to a log

›

update_nutrition_itemEdit a line item

›

delete_nutrition_itemRemove a line item

›

CHECK-INS · mood · energy · sleep · hydration · gut · skin · hangover (1–10)

get_checkinsRead scores

›

record_checkinSubmit today's scores

›

update_checkinEdit a check-in

›

delete_checkinRemove a check-in

›

FOOD SEARCH

search_foodName lookup

›

get_food_detailsFull macros

›

barcode_lookupUPC / EAN

›

get_favoritesSaved items

›

add_favoritePin to favorites

›

get_recentsRecently logged

›

SAVED MEALS

get_saved_mealsLibrary

›

create_saved_mealPin a combo

›

log_saved_mealRe-log in one shot

›

update_saved_mealEdit components

›

delete_saved_mealRemove from library

›

AI FEATURES

ai_chatCoach conversation

›

get_meal_suggestionsGoal-aware ideas

›

analyze_photoVision → macros

›

parse_food_descriptionText → structured

›

expand_meal_ingredientsBreak into parts

›

get_daily_greetingMorning summary

›

get_token_usageAI credits

›

CHAT THREADS

list_chat_threadsAll conversations

›

get_chat_threadOne thread

›

create_chat_threadNew conversation

›

delete_chat_threadRemove thread

›

generate_thread_titleAuto-name

›

get_chat_historyFull transcript

›

GOALS · daily_fiber_grams · daily_sugar_grams · daily_water_grams · daily_caffeine_mg · daily_alcohol_units

get_current_goalActive targets

›

list_goalsHistory

›

create_goalSet new target

›

update_goalEdit targets

›

delete_goalRemove goal

›

calculate_goalAuto-compute macros

›

WORKOUTS

get_recent_workoutsLast N sessions

›

list_workoutsDate range

›

create_workoutLog session

›

update_workoutEdit session

›

delete_workoutRemove session

›

NOTES & MEMORY

get_notesRead memory

›

create_noteSave a preference

›

update_noteEdit a note

›

delete_noteRemove a note

›

HOUSEHOLDS

get_householdsList households

›

get_household_summaryMember status

›

create_householdNew household

›

update_householdEdit details

›

delete_householdRemove household

›

list_household_membersAll members

›

suggest_meal_to_memberSend a recommendation

›

HOUSEHOLD INVITES

list_household_invitesAll invites sent

›

list_pending_invitesAwaiting response

›

create_household_inviteInvite a person

›

accept_inviteJoin a household

›

decline_inviteReject an invite

›

accept_invite_by_tokenVia link

›

GROCERY LISTS

create_grocery_listNew list

›

get_grocery_listsAll lists

›

get_grocery_listOne list

›

update_grocery_listEdit list

›

delete_grocery_listRemove list

›

add_grocery_itemsAdd items

›

check_grocery_itemsMark bought

›

uncheck_grocery_itemsUn-mark

›

set_item_pricesTrack cost

›

get_fridge_inventoryWhat's on hand

›

suggest_grocery_itemsGoal-aware ideas

›

60+ tools OAuth 2.0 + PKCE MCPB one-click npm global Source build macOS + Windows Node.js 20+ Claude Desktop

DataBody, inside Claude.