Skip to main content
Glama

Analyze lyrics for singing

lyrics
Read-onlyIdempotent

Analyzes lyric text for setting to music: each word's syllables, stress and phones (ARPAbet), the rhyme scheme with perfect and near rhymes, scansion against a poetic meter, and where each syllable can start on a tracker rhythm grid for a time signature (stressed syllables on strong beats, a pickup for an unstressed start), with warnings for lines that do not scan. EXPERIMENTAL: the singing voice and lyric setting are early features; word clarity and fit vary from song to song, so render and listen before relying on a result. It writes no melody and no words of its own (fit mode only reshapes a melody you send): the assistant writes the notes, then puts each syllable on its note in MTL (C4'twin- D4'kle; '_ holds a syllable over more notes). Word setting: mode=plan returns, for every line, exactly how many notes it needs and their lengths as a melody template (x:2 x:2 x:4: write the tune on that rhythm); mode=fit with melody checks and fits the words on a melody you wrote, adapting it (adapt, default true): extra syllables split a note (split=repeat repeats the pitch, pass steps toward the next note), a line that starts unstressed on the downbeat gets a pickup note, extra notes carry the syllable (melisma '_), and the last note is held to its bar line (hold); every change is listed, and warnings name syllables that are crammed, stress-inverted or off the beat and phrases that are shorter or longer than the bars (bars). Each fit line returns a ready MTL cell string and the adapted melody. Pronunciation comes from the dictionary; words guessed by rule are listed (fix with say). melody: one line per lyric line (a "\n" between lines, or an array). Tokens split by spaces: a pitch (C5, F#4, Bb3; octave 0-9, C4 = 48), optionally with its length in rows (C5:4; default = step rows); "." a rest of one step, ".3" three steps, ".3r" three rows; "x" a placeholder pitch (from plan mode: x:2); "@-2" first token only: the line starts at row -2 (a pickup before the bar; default 0). A note lasts until the next token. Example: "C5 D5 E5:4 . G5:2". Up to 48 lines per call. Schema a964948f.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "text": "<text>"
    },
    "name": "lyrics"
  }
}

Input Schema

NameRequiredDescriptionDefault
bpmNoTempo, for the shortest singable syllable (default 100)
rpbNoRows per beat (quarter note) of the pattern, default 4
sayNoPronunciations for names or invented words, ARPAbet with stress: {"kurtis":"K ER1 T IH0 S"}
barsNoBars per line; omitted, one count that suits every line
holdNofit: hold the last note of a line to its bar line (default true)
modeNoanalyze (default): grid rows only. plan: notes and rhythm each line needs (write the melody second). fit: put the words on melody (default when melody is sent)
stepNoRows per melody token without a length (default rpb/2, an eighth note)
textYesThe lyrics, one sung line per line
timeNoTime signature for the grid (default 4/4)
adaptNofit: change the melody to suit the words (split, pickup, melisma); false only reports (default true)
meterNoPoetic meter to scan against; auto picks the best fit
splitNofit: pitch of a note split in two: repeat it, or step toward the next note (default repeat)
detailNofull adds phones and meter notes to the text
melodyNoWith fit: one melody line per lyric line, e.g. "C5 D5 E5:4 . G5:2" (pitch[:rows], . rest, x placeholder, @-2 pickup start)
pickupNoAllow unstressed syllables before the bar (default true)

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / description
      Previous value: -"Schema 826121bd"New value: +"Schema a964948f"
  2. Changed7 schema fields changed
    • changedInput schema / description
      Previous value: -"Schema 2d544474"New value: +"Schema 826121bd"
    • addedInput schema / properties / adapt
      Added value: +{
      +  "description": "fit: change the melody to suit the words (split, pickup, melisma); false only reports (default true)",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / hold
      Added value: +{
      +  "description": "fit: hold the last note of a line to its bar line (default true)",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / melody
      Added value: +{
      +  "anyOf": [
      +    {
      +      "type": "string"
      +    },
      +    {
      +      "items": {
      +        "type": "string"
      +      },
      +      "type": "array"
      +    }
      +  ],
      +  "description": "With fit: one melody line per lyric line, e.g. \"C5 D5 E5:4 . G5:2\" (pitch[:rows], . rest, x placeholder, @-2 pickup start)"
      +}
    • addedInput schema / properties / mode
      Added value: +{
      +  "description": "analyze (default): grid rows only. plan: notes and rhythm each line needs (write the melody second). fit: put the words on melody (default when melody is sent)",
      +  "enum": [
      +    "analyze",
      +    "plan",
      +    "fit"
      +  ],
      +  "type": "string"
      +}
    • addedInput schema / properties / split
      Added value: +{
      +  "description": "fit: pitch of a note split in two: repeat it, or step toward the next note (default repeat)",
      +  "enum": [
      +    "repeat",
      +    "pass"
      +  ],
      +  "type": "string"
      +}
    • addedInput schema / properties / step
      Added value: +{
      +  "description": "Rows per melody token without a length (default rpb/2, an eighth note)",
      +  "maximum": 16,
      +  "minimum": 1,
      +  "type": "integer"
      +}
  3. Added

TDQS

A4.3/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnly/idempotent/non-destructive, and the description adds real context beyond them: an EXPERIMENTAL quality warning, the statement that it writes no melody or words of its own, that every fit change is listed, and a per-call limit of 48 lines. It stops short of full disclosure because it never describes the concrete return payload shape for analyze mode.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Purpose is front-loaded, but the bulk is one dense paragraph of semicolon- and parenthetical-chained clauses covering melody grammar, mode behavior and warnings. Most content earns its place, but a trailing 'Schema a964948f.' adds nothing and the lack of sectioning makes the mode/grammar explanation hard to parse on first read.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 15-parameter tool with no output schema, the description covers the mode semantics, the MTL melody grammar, output artifacts (ready MTL cell string, adapted melody, warnings, listed guessed pronunciations), and a call limit. Remaining gaps, such as the exact analyze-mode return or bpm/rpb/step interplay, are supplied by the schema, so the description is close to complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description meaningfully exceeds it: mode is explained in terms of what each returns, adapt is spelled out (split notes, pickup, melisma, held last note), split's repeat-vs-pass behavior is defined, bars warnings and say-based pronunciation fixes are explained, and the melody token grammar is documented. An agent gets parameter meaning here that the schema alone does not convey.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Analyzes lyric text for setting to music') and then enumerates the exact outputs: syllables, stress, ARPAbet phones, rhyme scheme with perfect/near rhymes, scansion against a meter, and grid placement. The three modes (analyze/plan/fit) are each defined, so an agent can tell this apart from render or song without opening another schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives a clear workflow: plan mode when the melody is not written yet, fit when a melody is sent, analyze as the default, plus an explicit 'render and listen before relying on a result' caveat. There is no explicit when-not-to-use or naming of a sibling tool as the alternative, so it stops short of a 5.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources