ClawHub
Back to skill

Security audit

do-issue

Security checks across malware telemetry and agentic risk

Overview

The skill mostly matches its issue-fixing purpose, but it also instructs agents to post tooling feedback to an external GitHub Discussions repo and can perform remote issue writes through broad triggers.

Install only if you are comfortable with an agent fetching issues, editing code, committing changes, opening a PR, commenting on issues, and possibly closing them. Before use, disable or ignore the external night-market feedback-posting step unless you explicitly want tooling observations sent there, and prefer narrow invocations with explicit issue numbers and review before any remote comments or closures.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The skill directs the agent to post workflow reflections to an external repository unrelated to the user’s current project. That creates an unnecessary outbound data channel and can leak repository names, branch state, workflow details, internal tooling behavior, or other sensitive context to a third party without explicit user approval. The surrounding wording normalizes this behavior as part of the standard workflow, which makes accidental disclosure more likely.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger list uses broad terms such as "github," "gitlab," "issues," and "automation," which can cause the skill to activate in contexts broader than intended. Because this skill can drive remote issue operations and spawn subagents, accidental invocation increases the chance of unintended repository actions, misleading task execution, or remote comments/closure workflows being initiated without clear user intent.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill documents commands for commenting on and closing remote issues, and later references issue updates, but it does not prominently warn users that invoking the skill may modify external systems. In a multi-agent automation context, omission of this warning is dangerous because users may trigger the skill expecting local analysis while the workflow proceeds to post comments or close issues on GitHub or GitLab.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.