ClawHub
Back to skill

Security audit

buildlog

Security checks across malware telemetry and agentic risk

Overview

This skill openly records and shares coding-session buildlogs, but users should treat uploads as potentially sensitive because they may include session text and file contents.

Install only if you are comfortable with a session-recording and sharing tool. Keep autoUpload off, consider setting buildlogs private and disabling file-content snapshots for private projects, and review recordings before uploading because they may include source code, prompts, logs, or secrets.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (4)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill encourages recording, exporting, and uploading AI coding sessions while the documented defaults enable public sharing and inclusion of file contents, but it does not present a clear up-front warning about the sensitivity of captured data. In this context, users may unintentionally expose source code, secrets, prompts, or prior conversation content through normal use of the skill.

Ssd 3

Medium
Confidence
91% confidence
Finding
The skill's core purpose is to capture and share coding sessions, so data export and upload are expected features; however, that same context makes accidental disclosure more likely because normal language commands can package and transmit sensitive interactions. Since coding sessions often include proprietary code, credentials, stack traces, and internal discussion, encouraging sharing without strong safeguards creates a real leakage path.

Ssd 3

Medium
Confidence
94% confidence
Finding
Plain-language commands like exporting the current session or the last N messages create an easy exfiltration mechanism for prior conversation content. In an AI coding environment, prior messages may contain secrets, internal code, or sensitive troubleshooting details, so the convenience of these commands increases the chance of unintentional disclosure.

Ssd 3

High
Confidence
98% confidence
Finding
The documented configuration sets `defaultPublic` to true and `includeFileContents` to true, which materially increases the chance that private code and session data will be exposed during ordinary use. Because the skill is specifically designed to record AI coding activity, these defaults are especially dangerous: they turn a logging feature into a likely data disclosure vector, even without malicious intent.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.