ai-compliance-review

• Outcome: Review
• Latest audit: Jul 25, 2026 · 8:32 AM UTC
• Version: 1.1.4

Overview

This compliance-review skill is cloud-based and purpose-aligned, but it sends reviewed text and an API key to a hard-coded plain-HTTP backend while the documentation does not clearly warn users or accurately match the shipped script.

Install only if you are comfortable sending reviewed text to the publisher's remote service. Do not submit confidential contracts, HR records, financial promotions, privacy policies, personal data, or regulated business material unless you have verified the backend, transport security, retention terms, and authorization to share that data. The shipped script and README-style commands also do not match, so expect setup and usage friction.

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• 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)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 87% confidence
• Finding: The skill declares environment and network capabilities via required API key usage and calls to an external DeepSeek endpoint, but does not declare corresponding permissions. This can undermine platform trust boundaries and informed consent because users may not realize the skill transmits content externally or accesses sensitive environment-provided secrets.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The usage section instructs users to submit text, files, and directory contents for review through a third-party API, but it does not place a clear warning at the point of use that sensitive legal, HR, contract, financial, or privacy-related data may leave the local environment. In this skill's context, that is especially risky because the advertised use cases commonly involve highly confidential business and personal information.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The `review` command transmits arbitrary user-supplied text to a third-party backend for processing, but the CLI does not provide a meaningful consent prompt or explicit warning at the point of submission. In this skill’s context, users are likely to submit sensitive marketing copy, contracts, hiring text, or privacy-related material, so silent transmission to a remote server creates a real confidentiality and privacy risk.

External Transmission

Medium

• Category: Data Exfiltration

Content

if method == "GET":
            resp = requests.get(url, headers=h, params=data or {}, timeout=30)
        else:
            resp = requests.post(url, headers=h, json=data or {}, timeout=60)
        if resp.status_code >= 400:
            try:
                err = resp.json()

• Confidence: 98% confidence
• Finding: The script sends data to an external API endpoint, including user review text and optionally an API key header, and the default backend uses plain HTTP to a hard-coded IP address. This exposes submitted content and credentials to interception or tampering in transit and increases the risk of data exfiltration to an untrusted remote service.

Static analysis

No suspicious patterns detected.