Security Audit

browserless-automation

github.com/ComposioHQ/awesome-claude-skills

AI SkillCommit 27904475d127

TRUSTED

Scanned about 2 months ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

browserless-automation received a trust score of 84/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.

SkillShield's automated analysis identified 2 findings: 0 critical, 1 high, 1 medium, and 0 low severity. Key findings include Broad execution capabilities via RUBE_REMOTE_WORKBENCH and RUBE_MULTI_EXECUTE_TOOL, Potential for data exfiltration through web automation.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.

Last analyzed on February 20, 2026 (commit 27904475). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

78%

Behavioral Risk Signals

Network Access

2 findings

Filesystem Write

1 finding

Shell Execution

2 findings

Dynamic Code

2 findings

Excessive Permissions

1 finding

Security Findings2

Severity	Finding	Layer	Location
HIGH	Broad execution capabilities via RUBE_REMOTE_WORKBENCH and RUBE_MULTI_EXECUTE_TOOL The skill instructs the LLM to use `RUBE_REMOTE_WORKBENCH` with `run_composio_tool()` for 'Bulk ops' and `RUBE_MULTI_EXECUTE_TOOL` for general tool execution. The term 'workbench' implies a powerful, potentially unconstrained execution environment. `RUBE_MULTI_EXECUTE_TOOL` allows the LLM to execute any dynamically discovered tool with arbitrary, schema-compliant arguments. If the underlying Composio tools or the workbench environment have broad system access (e.g., file system, network, arbitrary command execution), this grants excessive permissions to the LLM, creating a significant attack surface for malicious prompts to perform unauthorized actions or command injection. Implement strict sandboxing and capability-based security for `RUBE_REMOTE_WORKBENCH` and all tools executable via `RUBE_MULTI_EXECUTE_TOOL`. Ensure that tools can only access resources strictly necessary for their stated purpose. Explicitly define and limit the scope of operations for `run_composio_tool()`. Consider a whitelist of allowed tool slugs and arguments, and validate all inputs rigorously.	LLM	SKILL.md:50
MEDIUM	Potential for data exfiltration through web automation The skill's primary purpose is 'Browserless Automation', which inherently involves interacting with and potentially extracting data from web pages. While the skill itself does not explicitly exfiltrate data, the exposed powerful tools (`RUBE_MULTI_EXECUTE_TOOL`, `RUBE_REMOTE_WORKBENCH`) could be manipulated via prompt injection to extract sensitive information from browsed web pages and then transmit it to an external, unauthorized destination using other available Rube tools or network capabilities. Implement data loss prevention (DLP) mechanisms to monitor and restrict the outbound transfer of sensitive data. Ensure that any data extracted by Browserless tools is handled securely and only transmitted to authorized endpoints. Provide clear guidelines to the LLM on handling sensitive information and restrict its ability to send data to arbitrary external services.	LLM	SKILL.md:1

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/253886342b2bf655.svg)](https://skillshield.io/report/253886342b2bf655)