Analyzing Coverage

A user built a workout: bench press, overhead press, a couple of push-downs. Is it balanced, or is it all push and no pull? POST /analyze/coverage answers that for any set of exercises, in one call.

It's stateless — you send a list of exercise ids and get back an analysis. Nothing is stored. Send the workout a user is looking at right now and describe it back to them.

The call

Post the exercise ids as a JSON body.

SDK

import { createClient } from '@exercisedb/sdk';
const client = createClient({ apiKey: process.env.EXERCISEDB_API_KEY });

const coverage = await client.analyze.coverage([benchId, pullUpId]);

curl

curl -X POST 'https://api.harshitbishnoi.dev/analyze/coverage' \
  -H "x-api-key: exdb_…" \
  -H 'content-type: application/json' \
  -d '{ "exerciseIds": ["7b1088e6-…", "a2c9f1b0-…"] }'

The report describes what the set trains:

{
  "success": true,
  "data": {
    "exerciseCount": 2,
    "unknownExerciseIds": [],
    "muscles": {
      "primary": [
        { "slug": "chest", "count": 1 },
        { "slug": "lats", "count": 1 }
      ],
      "secondary": [
        { "slug": "front-delts", "count": 1 },
        { "slug": "biceps", "count": 1 },
        { "slug": "triceps", "count": 1 }
      ]
    },
    "balance": {
      "movementPatterns": [
        { "slug": "push", "count": 1 },
        { "slug": "pull", "count": 1 }
      ],
      "pushCount": 1,
      "pullCount": 1,
      "primaryMuscleGroupCount": 2
    }
  }
}

Reading the report

• muscles.primary / muscles.secondary — which muscles the set trains, and in how many of the exercises. A muscle appearing under primary with a high count is heavily worked; a muscle absent from primary entirely is a gap. Counts are sorted most-trained first.
• balance.movementPatterns — how the set distributes across the eight movement patterns.
• balance.pushCount / pullCount — the classic balance check. push counts push and squat patterns; pull counts pull and hinge. A workout with pushCount: 5, pullCount: 0 is the imbalance this field exists to surface.
• primaryMuscleGroupCount — how many distinct muscles the set hits as a primary mover. A low number over many exercises means the workout is narrow.

The report is descriptive, not prescriptive. It tells you what is and isn't covered; it doesn't tell the user what to add. That's a deliberate choice — you know your app's programming logic; the API gives you the honest picture to build on.

Unknown ids are reported, not fatal

If an id doesn't resolve to a real exercise — a typo, a malformed id, a stale reference — it lands in unknownExerciseIds and the analysis proceeds on the rest:

{
  "exerciseCount": 2,
  "unknownExerciseIds": ["not-a-real-id"],
  "…": "…"
}

Every id unknown is a 404

If none of the ids resolve, the endpoint returns 404 with code NO_EXERCISES_FOUND — there's nothing to analyze. As long as at least one id is valid, you get a report plus the unknowns list.

Limits

Send between 1 and 50 exercise ids. An empty list or more than 50 returns 400 with code VALIDATION_ERROR. Fifty covers any realistic single workout; for analyzing a whole program, call once per session.