ClawHub
Back to skill

Security audit

Coyns Wallet

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Coyns wallet skill, but it gives an agent broad authority to move virtual funds and use messages without clear confirmation limits.

Review before installing. Use a dedicated Coyns key with limited funds, treat COYNS_PRIVATE_KEY like money-control authority, and require explicit confirmation before any payment, exchange, deal creation, deal acceptance, deal completion, funding request, reward claim, registration, or message send.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The manifest uses very broad trigger language such as any mention of Coyns, wallet, or Playce.io currency, which can cause the skill to activate in contexts that do not clearly request a financial action. In a wallet/payment skill, over-invocation is dangerous because it increases the chance of unintended high-risk operations like transfers, exchanges, funding requests, or inbox access being proposed or executed without sufficiently specific user intent.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill advertises payment, exchange, escrow, funding, messaging, and inbox-reading capabilities without an explicit warning that these actions can move funds, commit assets to escrow, contact third parties, or expose private messages. In a financial skill, omission of side-effect warnings materially increases the risk of unsafe user consent and accidental execution of irreversible or sensitive operations.

Static analysis

No suspicious patterns detected.