ClawHub
Back to skill

Security audit

meta-social-collaboration-mind

Security checks for vulnerabilities and agentic risk

Overview

The skill is not overtly malicious, but it overstates its verification abilities and defaults to persistent local profiling without clear user controls.

Review this skill carefully before installing. It appears to run locally and does not show exfiltration or destructive behavior, but it should not be treated as a real self-verifying or self-evolving safety layer, and users should be comfortable with local cross-session storage of usage notes and preferences before enabling it.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill instructs use of `scripts/learner.py` to record outcomes, preferences, and reflections into local files such as `learned_patterns.json`, which is a file-write capability. Because no permissions are declared, operators and users are not clearly informed that the skill persists data to disk, weakening sandboxing and review controls. In this context the danger is elevated by the persistence of cross-session state and preference data, though the behavior appears local rather than overtly destructive.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The skill markets itself as a self-verifying, self-reflective, continuously evolving meta-skill with super-agent orchestration, but the described mechanics do not substantiate those claims and the analysis indicates only simple local recording and rule-based behavior. This mismatch is dangerous because users may overtrust outputs, assume safety checks occurred when they did not, and allow the skill to influence decisions under false assurances of reliability. The risk is amplified by repeated claims of enhanced verification and evolution, which can mask ordinary or immature behavior.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill states that after each use it automatically records usage outcomes and user preferences, but it provides no privacy notice, retention policy, access control description, or consent mechanism. This creates a privacy risk because personal preferences, error notes, and behavioral history may be silently accumulated across sessions and later reused in ways the user did not expect. The skill context makes this more concerning because the stored data is framed as automatic and persistent rather than optional and task-scoped.

Ssd 3

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs persistent collection and reuse of user preferences and operational history across sessions via `learned_patterns.json`. Persistent profiling without strong disclosure, minimization, or deletion controls can expose sensitive behavioral patterns and create secondary-use risk if the file is accessed by other components or users. In this skill, the danger is heightened because the persistence is positioned as a default self-improvement mechanism, encouraging continuous accumulation.

Static analysis

No suspicious patterns detected.