Skillv2.0.1

ClawScan security

Podcast Notes · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 18, 2026, 10:49 AM

Verdict: benign
Confidence: medium
Model: gpt-5-mini
Summary: This is a coherent, local bash-based podcast notes CLI that stores plain-text logs in your home directory and does not request credentials or network access; my confidence is medium because the provided script output was truncated so I couldn't inspect the very end of the file.
Guidance: This skill appears to do what it says: a local CLI that stores timestamped plain-text logs in ~/.local/share/podcast-notes and can export them to export.json/csv/txt. Before installing or running: 1) Be aware all inputs are stored unencrypted in your home directory—do not paste secrets or personal data you wouldn't want persisted. 2) Exports are written to the same data directory; review exported files before sharing. 3) The script uses standard tools (grep, date, wc); search uses grep with the search term you provide (grep will interpret the pattern), and JSON export appears to write raw values (may not escape quotes), so exported data could be malformed if entries contain arbitrary quotes/newlines—inspect exports if correctness matters. 4) The repository includes the bash script but no installer; installing typically requires copying the script into your PATH. 5) I inspected the provided files but the script output you supplied was truncated at the end, so I couldn't verify the final lines; if you want higher confidence, provide the complete script for review. Overall, there is no evidence of network calls or credential access and the behavior is consistent with the stated purpose.

Review Dimensions

Purpose & Capability: okThe name/description (podcast outlines, show notes, hooks, hashtags, exports) match the included script and SKILL.md. Required tools (bash, common Unix utilities) are appropriate and proportional; no unrelated credentials or binaries are requested.
Instruction Scope: okSKILL.md and the script limit activity to reading/writing log files under ~/.local/share/podcast-notes, showing stats, searching logs, and exporting data. There are no instructions to read unrelated system files, access remote endpoints, or exfiltrate data.
Install Mechanism: okNo install spec is provided (instruction-only), and the shipped implementation is a plain bash script. That is low-risk compared with downloading/executing remote payloads. Note: the repository includes the script file but does not provide an automated installer.
Credentials: okThe skill requests no environment variables or credentials. It uses $HOME to create a per-user data directory, which is expected. There are no unrelated secret or config path requirements.
Persistence & Privilege: okThe skill is user-invocable only (always:false) and does not request elevated privileges or modify other skills or system configs. It writes only to a per-user data directory.