Jingle maker

Runs the bundled nanoodle workflow workflows/jingle.noodle-graph.json against the NanoGPT API: a brief (Brief) feeds a lyric-writing LLM; the lyrics feed both a style-inventing LLM and a music model, which renders the sung track (Jingle, ~30 seconds by default). Requires Node.js 22.23.4+, 24.22.1+, or 26.12.1+ and the nanoodle npm package (npx nanoodle fetches it).

API key

The run needs a NanoGPT API key. Prefer NANOGPT_API_KEY already in the environment. Otherwise pass --env-file <path-to-.env> (this CLI: --env-file overrides ambient env).

Never print the key.

Run

From this skill's directory (or prefix paths if running from elsewhere). Replace the Brief value with the user's brief:

npx nanoodle run workflows/jingle.noodle-graph.json \
  --input "Brief=An upbeat 30-second jingle for a neighborhood coffee shop called Morning Ember. Warm, catchy, repeat the shop name in the hook." \
  --out ./jingle-out

Add --env-file .env only when the key is not already exported. Add --json for a structured result (paths, costUsd, remainingBalance). Music generation can take a minute or two — don't kill the run early.

Optional overrides:

--input "Lyric writer=<custom lyric-writer instructions>"   # the lyric LLM
--input "Style writer=<custom style-writer instructions>"   # the style LLM
--set "n4.duration=60"                                      # longer track (seconds)
--set "n4.instrumental=true"                                # no vocals

Inspect the interface anytime with:

npx nanoodle inspect workflows/jingle.noodle-graph.json

Inputs

The two optional keys are the node names shown in the editor. Each one replaces the system prompt of the node with that name.

Outputs

• Media is saved under --out as Jingle.<ext> — extension follows the audio MIME (often mp3). Use the path the CLI prints on the Jingle: line; do not hard-code the extension.
• With --json, the path is in outputs.Jingle.

Cost

Each run costs about $0.05–0.10 in NanoGPT credit (the Minimax-Music-02 song step dominates; the two LLM steps are fractions of a cent). Runs spend real credit — tell the user before running repeatedly.

If a run times out or is interrupted after submission, the remote job may still be running. Check its status and retain successful outputs before deciding to submit again. A fresh CLI invocation is not a durable resume and may bill again; do not automatically retry the whole workflow.