ClawHub
Back to skill

Security audit

Leadpoet

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign with medium confidence: the scans found no AIG issues and one low-probability concern that was judged inconsistent with the skill’s stated consent guidance, while some reference checks were incomplete.

The scanners support a benign verdict, but confidence is medium and Skillspector coverage was partial: three local references could not be matched to bundled files. Consider whether you are comfortable installing with those references unresolved; the scan does not establish what they contain.

A.I.G

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 104)May include surrounding context.

md
Pagination differs by operation. Follow its declared page/cursor contract;
bounded snapshots and `results_limited` are not complete exports. Do not infer
that a missing item does not exist. Never bypass consent, ownership, suppression,
credit, evidence or feature gates. Account consent remains in the application.