Step Audiobook

v0.1.4

Use when your agent needs to build, maintain, or run the local `audiobook` skill for voice-library management, Step official voice sync, clone voice analysis...

1· 75·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description target audio library, voice analysis, TTS and cloning workflows; required binaries (ffmpeg, ffprobe), required env var STEP_API_KEY, and included scripts (sync, analyze, build_tts_requests, clone helper, run_audiobook) all align with that purpose.
Instruction Scope
SKILL.md explicitly instructs the agent to read/write files under a local workspace (~/.openclaw/workspace), run the included Python scripts, and call Step endpoints (step_plan, step-audio-r1.1, stepaudio-2.5-tts). This behavior is expected for a local workflow tool, but it does mean user data (audio and text) and generated artifacts will be written to disk and (by default) sent to Step endpoints — the README warns about billing and requires explicit confirmation for paid clone actions.
Install Mechanism
Install spec only offers a brew formula to install ffmpeg/ffprobe (well-known package); no arbitrary downloads or extract/install of unknown binaries in the manifest.
Credentials
The skill requires a single credential STEP_API_KEY as primaryEnv and documents that this key is used for Step/LLM calls. No unrelated secrets or broad system credentials are requested. The code and templates reference only workspace paths under ~/.openclaw by default.
Persistence & Privilege
always:false and no indicators that the skill attempts to modify other skills or system-wide settings. It writes artifacts to its own workspace paths, which is normal for a workflow skill.
Assessment
This skill appears internally consistent with its stated purpose, but it performs network calls and writes local files — consider these before installing: 1) Use a test or limited STEP_API_KEY; the skill will call Step endpoints and may upload audio/text. 2) Paid cloning requires explicit confirmation (--confirm-paid-action); still double-check before running cloning commands. 3) Run initially in an isolated/sandbox workspace to verify behavior and review generated files under ~/.openclaw/workspace. 4) Inspect scripts if you have extra concerns about what is sent remotely (the repo documents endpoints and has a references/security.md). 5) Install ffmpeg/ffprobe from a trusted package source. 6) If you plan to change the LLM base_url, be aware that long-text reasoning calls will be sent to whatever base_url you configure (and will use STEP_API_KEY by default).

Like a lobster shell, security has layers — review code before you run it.

latestvk97cbhw08e4wr29f1sdtxh3n6s84sspa

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

🎧 Clawdis
Binsffmpeg, ffprobe
EnvSTEP_API_KEY
Primary envSTEP_API_KEY

Install

Install ffmpeg / ffprobe (brew)
Bins: ffmpeg, ffprobe
brew install ffmpeg

Comments