ClawHub
Back to skill

Security audit

PR Auto-Check

Security checks for vulnerabilities and agentic risk

Overview

The skill is a ClawHub maintainer workflow for pull requests; its GitHub access is disclosed and scoped, and the Discord-posting concern was not supported by the inspected artifact.

Install this only if you want an agent to help with ClawHub maintainer PR and issue workflows. Review any GitHub comments, labels, close actions, and proof-publishing commands before allowing them, because those actions affect the public repository.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill invokes shell scripts and external CLIs (`bash`, `gh`, `jq`, `curl`) but does not declare permissions or execution scope. That creates hidden capability for code execution and networked actions, making it harder for users or policy controls to understand that the skill can run commands and transmit data externally.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The trigger phrase `check this PR` is broad and likely to match routine conversation, which can cause unintended invocation of a workflow that runs scripts, inspects PR data, performs health checks, and posts results to Discord. Because the skill has side effects and external data transmission, accidental activation is more dangerous than for a read-only skill.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill description and workflow do not clearly warn that PR details, CI status, diff-derived review results, and health-check output will be sent to a Discord webhook. This can lead to unintentional exfiltration of potentially sensitive repository and operational information to an external third-party channel, especially if users assume the workflow is internal-only.

Static analysis

No suspicious patterns detected.