BestYou Coach Widgets

Render BestYou MCP tool responses as rich visual cards using OpenClaw's canvas.

First-Time Setup

Before calling any tools, check that mcporter is installed and the BestYou server is configured:

mcporter --config ~/.openclaw/workspace/config/mcporter.json list

• If mcporter is not found: npm install -g mcporter
• If bestyou is not listed: walk the user through setup. See references/setup.md for the full guide.

Setup summary:

1. User generates an API key in BestYou iOS (More → Connected Apps → OpenClaw)
2. Create config/mcporter.json with the key in the Authorization Bearer header (see references/setup.md for the exact JSON format)
3. Restart the OpenClaw gateway

Calling Tools via mcporter

All tools use this syntax:

mcporter --config config/mcporter.json call bestyou.\x3Ctool_name> [param=value ...]

If a custom config path is needed (e.g. calling from outside the workspace), use the full path:

mcporter --config ~/.openclaw/workspace/config/mcporter.json call bestyou.\x3Ctool_name> [param=value ...]

Examples:

mcporter call bestyou.get_account_link_status
mcporter call bestyou.get_daily_briefing date=2026-03-15
mcporter call bestyou.get_todays_action_plan date=2026-03-15
mcporter call bestyou.get_progress_snapshot date=2026-03-15
mcporter call bestyou.get_weekly_summary weekEndDate=2026-03-15
mcporter call bestyou.generate_workout type=strength duration=20 equipment=bodyweight experienceLevel=Intermediate goal="General strength"
mcporter call bestyou.analyze_meal_text description="chicken burrito bowl" timestamp=2026-03-15T12:30:00-06:00

If a custom mcporter config path is needed, add --config \x3Cpath>.

Tool-to-Template Map

Rendering Workflow

1. Call the BestYou MCP tool via mcporter
2. Read the matching template from assets/
3. Replace the sample data in the HTML with the actual API response values (see data mapping below)
4. Inline the CSS: copy the contents of assets/shared.css into a \x3Cstyle> tag in the HTML \x3Chead> so the output is self-contained
5. Present the populated HTML via canvas(action="present") or canvas(action="a2ui_push")
6. Provide a brief text summary alongside the visual

Design System

All templates use the BestYou Dark Glass design system (assets/shared.css):

• Background: #0a0a0a, cards: #141414, borders: #1e1e1e
• Font: Inter (Google Fonts)
• Color coding: green (#4ade80) = good, yellow (#facc15) = moderate, orange (#fb923c) = warning, red (#f87171) = poor
• Score thresholds: ≥70 green, 40-69 yellow/orange, \x3C40 red

Handling Missing Data

API responses may contain null or missing fields. Rules:

• If a score field is null, hide the gauge/bar or show "—" instead of a number
• If an array is empty or missing, omit that section entirely (don't render empty containers)
• If totalCalories or calorieGoal is null, omit the calorie summary row
• Never show "null" or "undefined" in rendered HTML

Data Mapping by Template

Daily Briefing

API response structure (nested):

Action Plan

Block icon mapping by blockType:

• wakeUp → 😴 (recovery icon bg)
• breakfast → 🍳 (nutrition icon bg)
• lunch → 🥗 (nutrition icon bg)
• snack → 🥤 (nutrition icon bg, or 🥑 for second snack)
• dinner → 🍽️ (nutrition icon bg)
• workout / cardio / strength → 🏃 (fitness icon bg)
• goToBed → 🌙 (recovery icon bg)

Block status: completed → ✅, upcoming or pending → ⬜

Progress Snapshot

Weekly Summary

Meal Analysis

Workout

Each exercise within a block:

• imageUrl → \x3Cimg> src (CDN: cdn.bestyou.ai)
• name → h3
• sets, reps, weight, rpe, rest → .rx-pill spans
• notes → .exercise-notes
• targetMuscle, exerciseType → .exercise-cues

Conversation Patterns

Morning check-in

User: "What's my day look like?" or "Morning briefing"

1. get_account_link_status → verify connected
2. get_daily_briefing → render briefing widget
3. get_todays_action_plan → render action plan widget
4. Summarize: readiness score, day type, top priority

Progress check

User: "How am I doing?" or "Show my progress"

1. get_progress_snapshot → render snapshot widget
2. Highlight strongest and weakest domains, top recommendation

Meal logging

User: "I just had [food description]"

1. analyze_meal_text with the description → render meal widget
2. Note protein hit vs target, suggest next meal bias

Workout request

User: "Give me a workout" or "[duration] [type] workout"

1. generate_workout with parameters → render workout widget
2. Brief summary: duration, focus, number of exercises

Weekly review

User: "Weekly summary" or "How was my week?"

1. get_weekly_summary → render weekly widget
2. Highlight trends (improving/declining), top achievement, next week focus

Text Summary Style

Keep text summaries alongside widgets brief and actionable:

• Lead with the key number (readiness score, calories, overall score)
• One sentence on what's going well
• One sentence on the top action item
• No need to repeat everything the widget shows visually