Back to skill

Security audit

Webapp Testing

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Playwright webapp testing helper, but users should treat its server-launch helper as trusted local command execution.

Install only if you are comfortable with a testing helper that can run local commands you provide, such as npm run dev or backend startup scripts. Review the exact --server commands, use it only in trusted workspaces, and check for leftover server processes after runs.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/with_server.py:69
Finding

Server Child Processes May Survive Lifecycle Cleanup

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The declared description frames the skill as Playwright-based webapp testing, but the documented behavior includes starting arbitrary shell commands, waiting on ports, and executing follow-up commands. That mismatch is dangerous because users or orchestration systems may trust the skill as a limited browser-testing utility while it actually has broader process-execution capability that could be abused to run unintended local commands.

Content

No source excerpt is available for this finding.

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
99% confidence
Finding

Because the tool's purpose is to orchestrate local webapp testing, it is likely to be run in developer environments with access to source code, secrets, SSH keys, and browser session data; this context increases the blast radius of shell-based parameter injection. A malicious project, test configuration, or copied command line could abuse --server to run arbitrary payloads while appearing to start a dev server.

Content

Scanner excerpt · scripts/with_server.py (reported line 69)May include surrounding context.

python
print(f"Starting server {i+1}/{len(servers)}: {server['cmd']}")

            # Use shell=True to support commands with cd and &&
            process = subprocess.Popen(
                server['cmd'],
                shell=True,
                stdout=subprocess.PIPE,

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
99% confidence
Finding

Because the tool's purpose is to orchestrate local webapp testing, it is likely to be run in developer environments with access to source code, secrets, SSH keys, and browser session data; this context increases the blast radius of shell-based parameter injection. A malicious project, test configuration, or copied command line could abuse --server to run arbitrary payloads while appearing to start a dev server.

Content

Scanner excerpt · scripts/with_server.py (reported line 69)May include surrounding context.

python
print(f"Starting server {i+1}/{len(servers)}: {server['cmd']}")

            # Use shell=True to support commands with cd and &&
            process = subprocess.Popen(
                server['cmd'],
                shell=True,
                stdout=subprocess.PIPE,

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill exposes shell execution and file-writing capabilities without declaring any explicit tool scope or permission boundaries. In an agent environment, this creates ambiguity about what the skill is allowed to do and increases the chance that a caller will invoke powerful operations without adequate review or sandboxing.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
98% confidence
Finding

The script passes a caller-controlled string directly to subprocess.Popen with shell=True, allowing shell metacharacters, command substitution, and chaining operators to be interpreted. If an attacker can influence the --server value, they can execute arbitrary OS commands under the privileges of the script, which is especially risky in an automation/testing tool that routinely launches local services.

Content

Scanner excerpt · scripts/with_server.py (reported line 69)May include surrounding context.

python
print(f"Starting server {i+1}/{len(servers)}: {server['cmd']}")

            # Use shell=True to support commands with cd and &&
            process = subprocess.Popen(
                server['cmd'],
                shell=True,
                stdout=subprocess.PIPE,

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/with_server.py (reported line 88)May include surrounding context.

python
# Run the command
        print(f"Running: {' '.join(args.command)}\n")
        result = subprocess.run(args.command)
        sys.exit(result.returncode)

    finally:

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The script saves screenshots to /mnt/user-data/outputs at L17 and L29, which is a file-write operation affecting user-accessible storage. While comments describe taking screenshots, there is no user-facing prompt, warning, or disclosure that image files will be created.

Content

No source excerpt is available for this finding.