AI Background Music Generator

The Specific Problem to Solve

In content creation workflows, background music (BGM) is crucial for enhancing immersion in videos, games, or podcasts. Traditional methods for obtaining BGM have notable bottlenecks: searching music libraries is time-consuming with copyright risks, custom composition requires professional skills or external collaboration, and existing AI tools often operate in isolation, hindering integration into automated development or Agent workflows. This makes music generation a efficiency barrier during bulk content production, especially for technically-inclined independent developers or small teams.

How the Skill Works

This skill provides AI background music generation via the aimv command-line tool (CLI), with the core functionality calling Hainian Music's cloud AI model to convert text descriptions into instrumental tracks. Its operation revolves around three key steps:

1. Initialization and Authentication: On first use, run node /bin/aimv.js init; the CLI outputs a login page link (e.g., https://lexuan.club/skills/connect?slug=bgm-maker&from=skill). After the user logs in via the web, they copy the API Key and run init --api-key <key> to complete the connection. This Key is long-term and syncs with the Hainian Music account.
2. Music Generation: Use commands like aimv bgm create or directly execute shell commands, inputting structured descriptions (e.g., "light electronic BGM for a city night vlog"). The AI generates two candidate tracks, with optional support for uploading reference audio to guide style. Generation is asynchronous, and the CLI provides status queries (aimv session tail) and error handling (e.g., wait_timeout indicates background generation in progress).
3. Result Management and Download: Once generated, works automatically sync to the Hainian Music account, allowing downloads via CLI or further editing in the mini-program 海绵音乐AI写歌 and web lexuan.club. Error tags (e.g., insufficient_points) directly guide users to recharge pages, creating a closed-loop operation.

The skill is designed for deep integration with development environments: when the Agent has shell execution permissions, the entire process can be automated via CLI; in restricted sandboxes, setting the environment variable AIMV_FORCE_CURL=1 forces the use of curl as the transport layer, or REST APIs can be called directly. This flexibility makes it suitable for embedding into CI/CD pipelines or batch processing scripts.

Applicable Boundaries and Considerations

• Environment Dependencies: Generation relies on Hainian Music's cloud services, requiring a stable internet connection; the local CLI needs a Node.js runtime, and the configuration directory can be redirected via the AIMV_CONFIG_DIR environment variable.
• Quotas and Permissions: Free quotas are limited; exceeding them returns API code=402 requiring recharge. Memberships can increase quotas. API Keys must be securely stored to prevent leaks.
• Generation Limits: Output is instrumental BGM, suitable for short looping tracks; it does not support vocal synthesis or complex arrangements. Description quality directly impacts results, requiring specific mentions of mood (e.g., "calm"), tempo (e.g., "medium pace"), and instruments (e.g., "piano").
• Workflow Boundaries: The skill focuses solely on music generation, not video editing or content integration; generated results require users to manage copyright independently, especially for commercial use.

This tool transforms AI music generation into a programmable interface, ideal for technical creators in rapid prototyping, content farm bulk scoring, or multimedia project automation, while noting its service-based nature and output granularity.