acestep-lyrics-transcription

SuspiciousAudited by ClawScan on May 10, 2026.

Overview

The skill mostly matches its transcription purpose, but its credential setup can expose API keys in logs and its inline Python path handling could run unexpected code with crafted paths.

Review or patch before use. Do not paste a production OpenAI/ElevenLabs key into this workflow unless config --set is changed to mask secrets; use a low-quota key and rotate it if exposed. Verify the configured provider/api_url, and avoid sensitive audio or crafted filenames/output paths until path handling is fixed.

Findings (4)

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.

What this means

An OpenAI or ElevenLabs key could appear in the conversation, terminal output, or logs. Anyone with access to that output could reuse the key and potentially consume quota or create charges.

Why it was flagged

The documented setup sets openai.api_key or elevenlabs.api_key through this config function; echoing $value will print the full API key to terminal or agent logs instead of masking it.

Skill content
mv "$tmp_file" "$CONFIG_FILE"; echo "Set $key = $value"
Recommendation

Patch the script to mask secrets when setting api_key fields, use a secure prompt or environment-based credential flow, declare the credential requirement in metadata, and rotate any key that was exposed.

ConcernMedium Confidence
ASI05: Unexpected Code Execution
What this means

A maliciously crafted filename or output path in a project could make the agent run unexpected local code during transcription or format conversion.

Why it was flagged

File paths are interpolated directly into Python source code passed to python -c. A crafted output path containing quotes or Python syntax could break out of the string and cause unintended Python execution.

Skill content
$PYTHON_CMD -c "... with open('$json_file', 'r', encoding='utf-8') ... with open('$output_file', 'w', encoding='utf-8') ..."
Recommendation

Pass paths as Python argv values, JSON-quote or otherwise safely escape them, and validate/normalize user-supplied output paths before invoking Python.

What this means

The audio content leaves the local machine and provider billing, retention, and privacy terms may apply.

Why it was flagged

The script uploads the selected audio file and sends the provider API key to the configured API endpoint. This is expected for a transcription integration, but it is sensitive behavior users should notice.

Skill content
-X POST "${api_url}/audio/transcriptions" ... -H "Authorization: Bearer ${api_key}" ... -F "file=@${audio_file}"
Recommendation

Use only trusted provider endpoints, verify api_url/provider settings, and avoid sending audio that you are not comfortable sharing with the selected transcription service.

What this means

Installation or first use may fail unexpectedly, and users may not realize from the registry metadata that a paid or quota-limited API key is needed.

Why it was flagged

The registry metadata under-declares requirements: SKILL.md and the script require curl, jq, Python, and an OpenAI or ElevenLabs API key.

Skill content
Required binaries (all must exist): none; Required env vars: none; Primary credential: none
Recommendation

Declare the required local tools and provider credential contract in metadata so users can make an informed install decision.