Back to skill

Security audit

ClawHub Security Scan

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly does what it claims, but it charges through a remote SkillPay API before running and has a scan-root isolation flaw users should review before installing.

Review this skill before installing. It performs a real paid SkillPay network charge before scan.py or precheck.py runs, so only use it if you accept the per-call billing, trust the SkillPay endpoint, and are comfortable providing a user ID and API key. Avoid scanning untrusted folders until symlink handling is fixed, because a crafted folder can make the scanner read files outside the intended scan root.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (2)

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
scripts/scan.py:281
Finding

Symlink Following Allows Reads Outside the Selected Scan Root

Content
View full analysis
List[ScanResult]: """Scan a single file for suspicious patterns""" results = [] seen_patterns: set[tuple[str, str]] = set() quoted_strings = re.compile(r'[\'"]([a-zA-Z0-9_\-]{16,})[\'"]') try: with open(file_path, 'r', encoding='utf-8', errors='ignore') as f: lines = f.readlines() ``` ```python def scan_directory(root_path: Path) -> List[ScanResult]: """Recursively scan all files in directory""" all_results = [] ignore_dirs = parse_gitignore(root_path) ignore_patterns = parse_security_config(root_path) for path in root_path.rglob('*'): if path.is_file(): # Check if filename is sensitive first sensitive = is_sensitive_filename(path.name) if sensitive: sensitive.pattern = "sensitive-filename" if sensitive.pattern not in ignore_patterns: all_results.append(sensitive) if should_scan_file(path, ignore_dirs): results = scan_file(path) ``` ### Technical Analysis The directory traversal treats every path for which `Path.is_file()` returns true as a file belonging to the scan root. File symlinks are followed by `is_file()` and later by `open()`. The implementation does not reject symlinks and does not compare the resolved candidate path against `root_path.resolve()`. Consequently, an untrusted directory can contain a symlink with a supported source extension, such as `external.py`, whose target is a readable file outside the directory selected by the user. The scanner will open and inspect that target using the privileges of the process running the scanner. This exceeds the minimum filesystem privile ...[truncated 2023 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
scripts/review.py:69
Finding

Untrusted Scan Report Fields Permit Terminal Escape-Sequence Injection

Content
View full analysis
Remediation
View remediation
str: text = str(value) text = ANSI_ESCAPE.sub('', text) text = ''.join( ch if ch in '\n\t' or ord(ch) >= 0x20 else repr(ch)[1:-1] for ch in text ) return text[:max_length] ``` 3. Apply sanitization independently to `file_path`, `line`, `description`, and `suggestion` before printing. 4. Restrict field lengths to prevent excessive terminal output and memory consumption. 5. Consider a safe rendering mode that uses escaped representations such as `unicode_escape` for reports from untrusted origins. 6. Add regression tests containing CSI sequences, OSC sequences, carriage returns, backspaces, embedded nulls, and oversized fields. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (29)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

A second description-behavior mismatch indicates the skill may access SKILLPAY_API_KEY, contact https://skillpay.me/api/v1, and charge users before running, none of which is plainly stated in the top-level purpose. Hidden credential use and billing-related network behavior create a trust and consent problem and could expose users to unexpected charges or secret handling.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

A second description-behavior mismatch indicates the skill may access SKILLPAY_API_KEY, contact https://skillpay.me/api/v1, and charge users before running, none of which is plainly stated in the top-level purpose. Hidden credential use and billing-related network behavior create a trust and consent problem and could expose users to unexpected charges or secret handling.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 30)May include surrounding context.

md
Scans for code patterns that commonly trigger ClawHub VirusTotal Code Insight suspicious flagging:

### 🔴 High Risk (really dangerous, should fix)
- Reads sensitive files (`/etc/passwd`, `~/.ssh/*`, `id_rsa`, etc.)
- Uses dangerous functions (`eval()`, `exec()`, `execfile()`) without validation
- Dynamic code execution from untrusted sources
- Hard-coded API keys/tokens in source code

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/precheck.py (reported line 56)May include surrounding context.

python
Scans for code patterns that commonly trigger ClawHub VirusTotal Code Insight suspicious flagging:

### 🔴 High Risk (really dangerous, should fix)
- Reads sensitive files (`/etc/passwd`, `~/.ssh/*`, `id_rsa`, etc.)
- Uses dangerous functions (`eval()`, `exec()`, `execfile()`) without validation
- Dynamic code execution from untrusted sources
- Hard-coded API keys/tokens in source code

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/review.py (reported line 39)May include surrounding context.

python
Scans for code patterns that commonly trigger ClawHub VirusTotal Code Insight suspicious flagging:

### 🔴 High Risk (really dangerous, should fix)
- Reads sensitive files (`/etc/passwd`, `~/.ssh/*`, `id_rsa`, etc.)
- Uses dangerous functions (`eval()`, `exec()`, `execfile()`) without validation
- Dynamic code execution from untrusted sources
- Hard-coded API keys/tokens in source code

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 62)May include surrounding context.

python
Scans for code patterns that commonly trigger ClawHub VirusTotal Code Insight suspicious flagging:

### 🔴 High Risk (really dangerous, should fix)
- Reads sensitive files (`/etc/passwd`, `~/.ssh/*`, `id_rsa`, etc.)
- Uses dangerous functions (`eval()`, `exec()`, `execfile()`) without validation
- Dynamic code execution from untrusted sources
- Hard-coded API keys/tokens in source code

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 333)May include surrounding context.

python
Scans for code patterns that commonly trigger ClawHub VirusTotal Code Insight suspicious flagging:

### 🔴 High Risk (really dangerous, should fix)
- Reads sensitive files (`/etc/passwd`, `~/.ssh/*`, `id_rsa`, etc.)
- Uses dangerous functions (`eval()`, `exec()`, `execfile()`) without validation
- Dynamic code execution from untrusted sources
- Hard-coded API keys/tokens in source code

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The script marketed as a pre-publish security check performs a mandatory external billing operation before any checklist logic runs. This creates a hidden side effect unrelated to the stated scanning purpose, sends user data to a third-party service, and can surprise users who expect an offline/local safety check.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
95% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/precheck.py (reported line 56)May include surrounding context.

python
It's okay to make API requests - this is what almost all skills do!

❌ DON'T:
  1. Read sensitive system files like /etc/passwd, ~/.ssh/*, /etc/shadow
     Skills almost never need this, it will definitely flag you as suspicious

  2. Use eval()/exec() on untrusted input

Credential Access

High
Category
Privilege Escalation
Confidence
95% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/review.py (reported line 39)May include surrounding context.

python
It's okay to make API requests - this is what almost all skills do!

❌ DON'T:
  1. Read sensitive system files like /etc/passwd, ~/.ssh/*, /etc/shadow
     Skills almost never need this, it will definitely flag you as suspicious

  2. Use eval()/exec() on untrusted input

Credential Access

High
Category
Privilege Escalation
Confidence
95% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 63)May include surrounding context.

python
It's okay to make API requests - this is what almost all skills do!

❌ DON'T:
  1. Read sensitive system files like /etc/passwd, ~/.ssh/*, /etc/shadow
     Skills almost never need this, it will definitely flag you as suspicious

  2. Use eval()/exec() on untrusted input

Credential Access

High
Category
Privilege Escalation
Confidence
95% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 333)May include surrounding context.

python
It's okay to make API requests - this is what almost all skills do!

❌ DON'T:
  1. Read sensitive system files like /etc/passwd, ~/.ssh/*, /etc/shadow
     Skills almost never need this, it will definitely flag you as suspicious

  2. Use eval()/exec() on untrusted input

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill is presented as a security scanner, but it also performs an undisclosed external billing action before doing any scan. This violates least surprise and creates a materially different capability than advertised: transmitting user identifiers and enforcing payment through a remote service.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 259)May include surrounding context.

python
# Sensitive filenames that should not be in your skill repo
SENSITIVE_FILENAMES = [
    '.env', '.env.*',
    '*.pem', '*.key', '*.cer', '*.crt',
    'id_rsa', 'id_dsa', 'id_ecdsa', 'id_ed25519',
    '*.secret', '*.token',

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 270)May include surrounding context.

python
# Sensitive filenames that should not be in your skill repo
SENSITIVE_FILENAMES = [
    '.env', '.env.*',
    '*.pem', '*.key', '*.cer', '*.crt',
    'id_rsa', 'id_dsa', 'id_ecdsa', 'id_ed25519',
    '*.secret', '*.token',

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · scripts/scan.py (reported line 270)May include surrounding context.

python
fn = filename.lower()
    
    sensitive_patterns = [
        (r'\.env', "Found .env file - may contain API keys/secrets"),
        (r'\.pem$', "Found PEM private key file"),
        (r'\.key$', "Found key file"),
        (r'id_', "Found SSH private key file"),

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The main execution path blocks scanning unless a remote payment request succeeds, even though billing is not necessary to the technical function of static code scanning. This creates hidden network dependency and payment-gating that users would not expect from the stated purpose of the tool.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module describes itself only as an interactive checklist wizard, but it also enforces billing before the checklist starts. This misleading description can cause users or reviewers to trust and run the script without realizing it performs a paid network transaction.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The billing function transmits user_id and uses an API key in a network request without any user-facing warning, consent, or dry-run mode. Even if the endpoint is legitimate, silent transmission of identifying and billing-related data is risky and inappropriate for a tool presented as a security checklist.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

Reading SKILLPAY_API_KEY from the environment is not inherently unsafe, but in this context it is unrelated to the advertised purpose of a pre-publish security scan and enables undisclosed credentialed network actions. That mismatch increases the risk of credential misuse and violates user expectations about what the tool does.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The script presents several user-facing status and error messages only in Chinese, while other prompts are in English. This imposes a locale/language choice on users without offering selection or documenting that the tool is intended only for a Chinese-speaking audience.

Content

No source excerpt is available for this finding.

Unrestricted Tool Access

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill grants unrestricted tool access without appropriate constraints. An agent with unfettered tool access can perform arbitrary actions including file modification, network requests, and code execution.

Content

Scanner excerpt · scripts/precheck.py (reported line 62)May include surrounding context.

python
2. Use eval()/exec() on untrusted input
     This is a critical security vulnerability, avoid if at all possible

  3. Download and execute arbitrary code from external sources
     If you must do this, only use HTTPS from trusted sources

  4. Exfiltrate user files, environment variables, or system information

Unrestricted Tool Access

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill grants unrestricted tool access without appropriate constraints. An agent with unfettered tool access can perform arbitrary actions including file modification, network requests, and code execution.

Content

Scanner excerpt · scripts/review.py (reported line 45)May include surrounding context.

python
2. Use eval()/exec() on untrusted input
     This is a critical security vulnerability, avoid if at all possible

  3. Download and execute arbitrary code from external sources
     If you must do this, only use HTTPS from trusted sources

  4. Exfiltrate user files, environment variables, or system information

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The module docstring claims the file only scans skill code for suspicious patterns, but the script also charges users via SkillPay. Misrepresenting behavior reduces transparency and can lead users to run code that performs unexpected external actions.

Content

No source excerpt is available for this finding.