openclaw audit trail

• Outcome: Pass
• Latest audit: May 29, 2026, 8:20 PM
• Version:
  1.0.0

Overview

This skill is a coherent local audit-log tool, but users should handle its logs and exports as sensitive records.

Install only if you are comfortable keeping durable local records of AI prompts, context, reasoning, decisions, and costs. Use a protected storage path, avoid logging secrets or unnecessary personal data, review exports before sharing them, and consider a pinned or project-local npm install after checking the package source.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The README explicitly encourages recording prompts, context, reasoning steps, outputs, and costs, which commonly contain personal data, secrets, internal business logic, or regulated information. Because this is an audit-trail skill intended for broad integration, omission of any warning, minimization guidance, or redaction controls can lead users to persist sensitive data in logs and later expose it through storage, queries, or exports.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The export examples promote generating JSON, HTML, and CSV reports and even include reasoning, but do not caution that exports are easy-to-share secondary copies that can leak sensitive prompts, personal data, or internal chain-of-thought-like content. In an audit product, export functionality materially increases exposure because data leaves the original storage boundary and may be opened in browsers, emailed, or imported into less secure systems.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill explicitly promotes recording prompts, reasoning, outputs, and context, yet provides no warning that these fields may contain PII, secrets, regulated data, or confidential business information. In an audit-trail tool, this omission is dangerous because users may enable comprehensive logging and export sensitive decision traces into local files or reports that persist longer than intended and are later exposed through backups, sharing, or weak filesystem controls.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The examples normalize logging detailed reasoning plus user-related context such as recipient email addresses, loan application attributes, moderation metadata, and operational decision traces without any privacy, compliance, or retention warning. This is more dangerous in context because the skill is specifically designed to preserve and export immutable audit records, which can turn sensitive model inputs and inferred reasoning into durable, searchable artifacts that may violate privacy expectations or compliance obligations.

VirusTotal

65/65 vendors flagged this skill as clean.