Voice Log
Analysis
This skill appears purpose-built for voice journaling, but it starts a background microphone recorder that sends audio to Soniox and stores transcripts, so it deserves careful review before installation.
Findings (6)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
detached: true,
stdio: ["ignore", outFd, errFd],
env: daemonEnv,
});
child.unref();Starting the journal launches a detached background process that continues independently of the initial command.
const { spawn, spawnSync } = require("child_process"); ... const result = spawnSync("which", [cmd], {The daemon invokes local binaries to locate and run microphone capture tools. This is expected for the purpose, but it relies on local PATH-resolved programs.
Source: unknown; Homepage: none
The registry metadata provides limited provenance for a skill that installs/runs Node code and accesses the microphone.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
Requires SONIOX_API_KEY ... Required env vars: none ... Primary credential: none
The skill requires a Soniox API key, but the registry requirement summary under-declares the credential. The key use is expected for Soniox STT, but it is still an account permission boundary.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
Captures microphone audio locally and streams audio to Soniox realtime STT only while journal is running.
The skill discloses a third-party provider data flow for live microphone audio. This is purpose-aligned, but live speech is highly sensitive.
Keeps a text-only log file, with live conversation logs bucketed by minute. Keeps only the latest 60 minutes
The skill persists transcribed conversation text and later returns it to the agent for summaries. This is disclosed and bounded, but it creates sensitive local context.
