midasheng-audio-generate
v1.1.5Generate immersive audio scenes—complete with speech, sound effects, music, and ambient sounds by text descriptions.
⭐ 1· 128·1 current·1 all-time
byJunbo Zhang@jimbozhang
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The skill claims to convert text into immersive audio and its runtime instructions perform exactly that: craft a structured prompt and POST it to a remote audio-generation API. The required functionality (prompt engineering + curl call to the service) matches the described purpose. Minor inconsistency: the registry metadata lists no required binaries while SKILL.md lists 'requirements: curl' (the curl command is used in the instructions).
Instruction Scope
Instructions are narrowly focused: they direct the agent to build a structured, lowercased prompt and send it to the specified API endpoint, and to optionally check a queue-status endpoint. The instructions do not ask the agent to read local files, other env vars, or system state. Important privacy note in SKILL.md: user-provided prompts are transmitted to an external endpoint and data retention is unknown — the skill explicitly warns not to include PII or sensitive content.
Install Mechanism
This is an instruction-only skill with no install spec and no code to write to disk, which is the lowest-risk install model. The only runtime requirement is using curl (per SKILL.md), which is a normal CLI tool for making HTTP requests.
Credentials
The skill does not request any environment variables, credentials, or config paths. That is proportional to its described purpose. Note: the SKILL.md declares the remote endpoint accepts unauthenticated requests (authentication: none); if the endpoint actually requires credentials, the skill might fail or prompt for extra setup not declared here.
Persistence & Privilege
The skill is not marked always:true and does not request persistent privileges or attempt to modify other skills or system-wide config. It can be invoked by the agent normally (default).
Assessment
This skill behaves like a thin adapter that sends your text to an external Xiaomi-hosted audio-generation service. Before installing: (1) Do not send PII or confidential text — SKILL.md explicitly warns data may be retained and is sent offsite; (2) verify you are comfortable with the endpoint host (llmplus.ai.xiaomi.com) and its privacy/retention policy — the skill does not document retention or authentication; (3) note the small metadata mismatch: registry metadata omitted 'curl' while SKILL.md requires it; (4) test with non-sensitive prompts first to confirm the service accepts unauthenticated requests and to observe latency/quality; (5) if you require stronger privacy, consider using a local or self-hosted alternative (the SKILL.md links include demos and a GitHub repo you can review). If you want higher assurance, ask the publisher for explicit data-retention and authentication details or prefer a skill with documented privacy guarantees.Like a lobster shell, security has layers — review code before you run it.
latestvk976a26grkhdfw04dthehwxcvh8393cw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.