Recipe catalog / mood-match

Check a passage against a requested mood

Does the described passage deliver requestedMood?

A listener asked for a mood and a passage has been generated or described, and you want to check the passage delivers that mood before playing or committing to it.

Explore this recipe interactively · Source and implementation guide

Use mood-match in TypeScript

Install with npm install jev-recipes. Requires Node.js 22.9 or newer and ES modules. Set TYPESAFE_API_KEY in your server environment for live calls, which send input to TypeSafe and use API quota. See the installation guide.

import { moodMatch } from 'jev-recipes/mood-match';

const result = await moodMatch({
  "requestedMood": "calm, something to relax to",
  "passage": "Fast tempo around 160 bpm, driving repeated sixteenth notes in the left hand, diminished and augmented chords with frequent chromatic shifts, sudden accents on off-beats, fortissimo throughout, melody leaping in the high register.",
  "minConfidence": 0.8
});
console.log(result);

Input contract

{
  "input": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "requestedMood": {
        "type": "string"
      },
      "passage": {
        "type": "string"
      },
      "minConfidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      }
    },
    "required": [
      "requestedMood",
      "passage"
    ]
  },
  "result": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "model": {
        "type": "string"
      },
      "usage": {
        "type": "object",
        "properties": {
          "input_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          },
          "output_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          }
        },
        "required": [
          "input_tokens",
          "output_tokens"
        ],
        "additionalProperties": false
      },
      "status": {
        "type": "string",
        "enum": [
          "ready",
          "review"
        ]
      },
      "probability": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      },
      "confidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      },
      "verdict": {
        "type": "string",
        "enum": [
          "matches",
          "mismatched"
        ]
      }
    },
    "required": [
      "model",
      "usage",
      "status",
      "probability",
      "confidence",
      "verdict"
    ],
    "additionalProperties": false
  }
}

Saved example result

This hand-authored response demonstrates the contract. It is not a model accuracy measurement. Run it without an API key: npx jev-recipes demo mood-match.

{
  "model": "demo-fixture",
  "usage": {
    "input_tokens": 0,
    "output_tokens": 0
  },
  "status": "ready",
  "probability": 0.04,
  "confidence": 0.96,
  "verdict": "mismatched"
}

Evaluation evidence

Use the evaluation guide to measure this decision on your own labeled cases.

Limitations

• Compares two texts as written. It cannot hear the passage and does not know whether the description matches what was played.
• Judges conventional musical affect, not the reaction of any particular listener, who may hear the same passage differently.
• A mismatched verdict flags the passage for a change; it does not say what to change.

Related recipes

• audience-fit: Use audience-fit to check material against a described audience rather than a described musical passage against a requested mood.
• instruction-fit: Use instruction-fit to check whether an instruction's scope covers a task, rather than whether music delivers a feeling.