JustGains | Workouts & Nutrition Tracking のリモート MCP サーバーとは何ですか?
JustGains | Workouts & Nutrition Tracking のリモート MCP サーバーは、https://justgains.com/connect/mcp にホストされた Model Context Protocol エンドポイントです。AI アシスタントはローカルに何もインストール・実行することなく接続できます。
JustGains | Workouts & Nutrition Tracking MCP Server認証済み注目
@Just Super Human Foundation
基本情報
JustGains | Workouts & Nutrition Tracking MCP Server について
JustGains | Workouts & Nutrition Tracking is a remote MCP server available at https://justgains.com.
接続情報
https://justgains.com/connect/mcpセットアップ
claude mcp add justgains-workouts-nutrition-tracking --transport http https://justgains.com/connect/mcpJustGains | Workouts & Nutrition Tracking MCP Server のツール
25Display a workout in the preview editor by its ID
Generate a workout based on a natural language description. Examples: "30 minute leg workout", "Upper body workout with dumbbells", "Full body HIIT circuit", "Beginner chest and triceps workout"
Get the signed-in user's training progress (daily gains, workout volume, steps, active days) over a timeframe. Provide startDate/endDate (YYYY-MM-DD) or a trailing number of days. Defaults to the last 30 days. Requires the user to have connected their JustGains account.
Search the JustGains exercise library by name (optionally filtered by muscle group or equipment). Returns exercise codes and supported measurements to use when building or updating a workout.
Create a workout in the signed-in user's JustGains account, saved under their own creator profile. Provide a title and an ordered list of exercises (by code from search_exercises, or by name to auto-resolve); each exercise can include sets with reps/weight/RPE. Use this for precise, user-specified workouts (use generate_workout for AI-generated ones).
Update an existing workout in the user's JustGains account. Provide the workoutId plus any of: new title, new description, or a replacement list of exercises. Call show_workout first if you need the current exercises before modifying them.
List the signed-in user's most recent workouts (title, date, duration, total volume, sets, reps, exercise count). Use to discuss recent training, consistency, and volume trends.
Get detailed information about a specific exercise by its code. Returns the exercise name, instructions, equipment, muscles worked, and measurements.
Get the CURRENT user's workout-log history (workouts they actually performed and recorded). Use this whenever the user asks about THEIR OWN workouts - 'my last workout', 'what did I do this week', 'how much did I bench Monday', 'show my recent sessions', 'workouts in March', 'last summer's training'. For any historical timeframe, pass `since` and/or `until` (ISO-8601). The tool will paginate deeper automatically until the requested window is covered (or `maxLogs` is reached). For RANKED questions ('what was my best workout', 'longest session ever', 'highest-volume day'), set `rankBy` ('volume' | 'duration') and `topN`. The tool then sweeps the window, computes the metric per log, and returns the top N logs sorted by that metric instead of newest-first. Returns logs with id, title, date, exercise count, and a per-exercise summary of sets/reps/weight; when ranking, each log also carries totalVolumeKg/totalVolumeDisplay and durationSeconds. Returns an error if the user is not signed in.
Get the CURRENT user's recorded history for a SINGLE exercise (by exerciseCode), plus their personal best. Use when the user asks 'show my bench press progression', 'how heavy have I squatted', 'what's my best deadlift', 'when did I last do pull-ups'. You usually want to call search_exercises FIRST to resolve a name to a code. For ANY HISTORICAL TIMEFRAME ('my bench in 2024', 'squats last summer'), pass `since` and/or `until` (ISO-8601). The tool sweeps deeper pages automatically and returns only sessions in the window: DO NOT page manually. Use `maxSessions` to cap the scan. Returns history items (workout id, completed-at, sets) sorted newest-first by default, plus a personalBest summary if any has been recorded. Requires sign-in.
Search the workout-program catalog. Returns paginated programs with title, description, total weeks/days, rating, creator, and a public link. Use when the user asks about available programs ('show me programs for hypertrophy', 'find a beginner program', 'what programs do you have'). Optional filters: search keyword, creatorId.
Get a workout program (template) by id, including its weeks and days. Use after search_workout_programs or get_my_workout_programs to dig into a specific program. Set includeData=false for a lightweight summary that omits per-day workout content.
Update the CURRENT user's enrollment on a workout program. Use to start, abandon, or complete a program: 'start this program', 'I'm done with PPL', 'abandon the cut program'. Requires the program id (use search_workout_programs to find it). Action must be one of: start | abandon | complete. Note: 'start' uses the Purchasing status, matching the existing API behavior.
Create a brand-new multi-week/multi-day workout PROGRAM for the user and save it to their account. You write the FULL program out as plain text in `programText`: a title, then each week and day with that day's exercises and their sets × reps (add weight or RPE when relevant). Rest days can be written as 'Rest'. The text is parsed into a structured program (real exercises are matched automatically) and saved. Use this whenever the user asks you to build / create / make / design a training PROGRAM, a multi-week or multi-day plan (PPL, full-body split, 5x5, a cut block, etc.). For a single one-off workout use generate_workout instead. If the user pasted their own program, pass it through verbatim.
Generate a real, road-and-path-snapped ROUTE (a plannable run / ride / walk / hike path) starting and ending at the user's location, and save it to their Routes tab. Use when the user asks you to build / make / plan a route, a running loop, a bike ride, etc. Use startLat/startLng from the app-data block's current location when present, or coordinates the user explicitly gives; NEVER guess them. If neither is available, OMIT startLat/startLng and the route starts from the user's last known training location (the start of their most recent GPS activity or saved route). Before calling, make sure you know the activity type and the distance they want (ask if they haven't said). The generated route follows actual streets and paths and lands close to the requested distance.
Get the CURRENT user's bodyweight log (entries plus summary: current, lowest, highest). Use whenever the user asks about their bodyweight: 'what do I weigh now', 'how much have I lost', 'show my weight trend', 'when did I last weigh in', 'what was my lightest'. For ANY HISTORICAL TIMEFRAME ('what did I weigh in March', 'last summer', 'between Jan 1 and Feb 14'), pass `since` and/or `until` (ISO-8601). Summary stats (current/lowest/highest) are returned for BOTH the full history and, when a window is set, the windowed slice. Optional `limit` truncates the entry list. Requires sign-in.
Record a new bodyweight entry for the CURRENT user. Pass `weightKg` (kilograms) OR `weightLbs` (pounds), the tool converts lbs → kg before storage. If the user spoke in pounds, prefer `weightLbs` so the round-trip is exact. `recordedAt` is optional ISO-8601 (defaults to now). Valid stored range: 25kg–500kg (≈55–1100 lbs); the API rejects values outside this.
Read the CURRENT user's food diary, daily totals, targets, local date/timezone and last 30 days of foods (up to 100 entries). Call when the user wants to see their diary, when a request depends on a past day ('same as yesterday'), or to find entries to correct or remove that no recent food result returned. Routine logging does not need it: log_my_food resolves the local date and checks recent entries on the backend. Never infer that a planned meal was eaten.
Search the cached food catalog and matching recent diary entries. This never waits on an external nutrition provider. Use only when the user asks to find or compare foods. For logging, call log_my_food directly with estimates instead. Results contain up to five identity-checked candidates and three possible variants; missing results do not prevent logging an estimate. Preserve brand and raw/cooked preparation.
Log foods the CURRENT user actually ate in one call. Call this directly: do not search_foods, browse, research nutrition, or read the diary first. Send every food/drink with quantityGrams and a quick estimatedMacros fallback for the whole portion; round estimates rather than researching exact numbers. The API checks recent foods and the cached catalog, prefers reliable nutrition, and labels estimate fallbacks. Preserve brand and raw/cooked preparation. Use servingQuantity/servingUnit for counts and portionEstimated=true for guessed weights. Briefly disclose assumptions; do not ask for missing flavor or grams. Use a new requestId per meal and keep it unchanged on retries. Follow-up corrections reuse diaryEntryId and expectedUpdatedAt from the last result, keep the serving count, and omit grams to recalculate a corrected catalog serving. Only read the diary for a past-day reference, an explicit repeat, or a correction/removal whose ID is unavailable. Use recentEntryId only for an explicit repeat. After success, give a short confirmation; do not call another tool to verify the save or repeat the card. Never log plans or hypothetical meals.
Remove foods the CURRENT user asks to delete from their diary, e.g. 'remove those crackers', 'I did not eat that', or 'delete breakfast'. Identify the exact saved entries from the user's request and the latest food result (log, removal or diary read); call get_my_food_diary for the relevant day only if no recent result has them. Pass their diaryEntryId and expectedUpdatedAt exactly as that result gives them. Remove only requested entries; ask which one if the target is ambiguous. For an entire meal, pass all entries in that meal, not the whole day. Removing one pack from a two-pack entry is a portion correction with log_my_food, not removal of the entire row. The API removes the batch atomically, updates day totals and syncs deletion to mobile. Repeating the same IDs is safe. This removes diary entries, never catalog foods. Do not call unless the user requested removal.
Fetch the CURRENT user's full profile snapshot: name, email, bio, measurement system, timezone, plus private meta (age/birth date, height, gender, experience level, fitness goals, injury tracking, baseline weight). Use whenever the user asks 'what's my profile', 'how old am I', 'what's my height', 'what experience level am I set as', 'what goals do I have'. Pair with `update_my_profile` if they want to change any of these.
Bundle a one-call snapshot of the CURRENT user's recent training: count of workout logs in the last `days` (default 7), per-muscle-group recovery data, weight delta vs `days` ago, and the most recent log id. Use for "how was my week", "quick recap", "how am I doing". For HISTORICAL questions ('how was my week ending June 1, 2024'), pass `endDate` (ISO-8601), the window becomes (endDate - days, endDate) instead of anchoring to today. Saves the agent from chaining get_my_workout_logs + get_my_recovery_analytics + get_my_weight_history.
Analyze the CURRENT user's workout consistency over a window: longest streak, current streak (today/yesterday inclusive), distinct training days, total logs, and average workouts per week. Use for 'how's my consistency', 'what's my streak', 'how often have I trained lately'. Default window 60 days, max 365. For HISTORICAL consistency questions ('how consistent was I in 2024'), pass `endDate` (ISO-8601), the window becomes (endDate: days, endDate). Current-streak only counts when the window ends today/yesterday.
Cross-exercise leaderboard for the CURRENT user. Use whenever the user asks about their TOP / MOST / HEAVIEST exercises across the board: 'what's my most used exercise', 'top 3 lifts', 'heaviest exercises', 'what do I lift the most'. Scans the user's workout logs (default 365 days, capped at `maxLogsScanned`) and returns three rankings in one call: byFrequency (session + set count), byMaxWeight (peak single-set weight), and byVolume (total tonnage). Each entry includes the exerciseCode + name so the agent can drill into a specific lift via get_my_exercise_history. For HISTORICAL questions ('most-used in 2024', 'top lifts last summer'), pass `since`/`until` (ISO-8601). Use `topN` to widen/narrow each list (default 5, max 25). Requires sign-in.
概要
JustGains is a workout and nutrition app. This is its official hosted MCP server: an agent can build and save workouts, read your training history, plan routes, and log meals and bodyweight in your JustGains account.
Endpoint: https://justgains.com/connect/mcp (Streamable HTTP)
What your agent can do
Workouts
- Generate a saved workout from a plain request like "30-minute dumbbell push day"
- Create or edit a workout with exact exercises, sets, reps, loads and set types (warmup, drop set, myo-rep, failure)
- Search the JustGains exercise library by name, muscle group or equipment
Training history and progress
- Completed workouts with dates, exercises and top sets
- Every recorded set of one lift, with your personal best
- Top exercises, weekly recap, streaks and training frequency
Programs and routes
- Browse, create, start or leave multi-week programs
- Plan a run, ride, walk or hike route snapped to real roads and paths
Food and bodyweight
- Log a whole meal in one call. When a portion is missing, the server saves an estimate and marks it as one
- Read your food diary and daily macros, correct an entry, or remove one
- Log bodyweight and see the trend
Sign-in
The server uses OAuth 2.0 with PKCE and dynamic client registration. An unauthenticated request gets a 401 challenge, so your client opens the JustGains sign-in on its own. You don't need an API key. Eneration work without anaccount. Everything else reads and writes the signed-in user's own data.
- Homepage: https://justgains.com
- Docs: https://justgains.com/learn/use-justgains-in-claude
- Repository: https://github.com/JustGains/justgains-connector
よくある質問
JustGains | Workouts & Nutrition Tracking の MCP サーバーにはどう接続しますか?
エンドポイント https://justgains.com/connect/mcp を Claude Code、Cursor、VS Code など任意の MCP 対応クライアントに追加してください。このページのセットアップ手順で各クライアントを一度に設定できます。
JustGains | Workouts & Nutrition Tracking の MCP サーバーは認証が必要ですか?
はい。JustGains | Workouts & Nutrition Tracking は OAuth を使用します。初回接続時に MCP クライアントがブラウザを開いてサインイン・認可を行い、以降はその認証情報を再利用します。
JustGains | Workouts & Nutrition Tracking の MCP サーバーはどのトランスポートを使用しますか?
JustGains | Workouts & Nutrition Tracking は Streamable HTTP エンドポイントを公開しており、これはリモート MCP サーバーで広く使われ、主要な MCP クライアントがサポートするトランスポートです。
他のリモート MCP サーバー
AIsa
AIsa is one key for 950+ data APIs across SEO and AI visibility, finance, social, web search, sales and agent mail — the go-to-market data stack behind prospecting, competitor research and brand monitoring. Five tools ra
AWS Knowledge MCP Server
Search and read current AWS documentation, check regional availability and load AWS agent skills, with no AWS account needed.
42min
Official 42min MCP server: let Claude, ChatGPT, Cursor or any MCP client manage your meeting scheduling. Check availability, find free slots, book, cancel and reschedule meetings, manage recurring series, secret booking
AQL PropertyCheck — Physical check of one exact address
Check one exact Gold Coast or Brisbane City property address. Free address/coverage check and short summary; paid structured Agent JSON adds available building, street and surrounding context, with Brisbane routed journe
AdLoop Cloud
Hosted AdLoop: Google Ads, Reddit Ads, GA4 and Tag Manager in claude.ai, ChatGPT and more, with a preview before every change.
コメント