ClawHub
Back to skill

Security audit

Coinbase Agent

Security checks for vulnerabilities and agentic risk

Overview

This skill is not deceptive, but it asks for autonomous control over live Coinbase trading and crypto transfers without enough user confirmation or safety scoping.

Review carefully before installing. Only use this with tightly scoped Coinbase API keys, disable withdrawals or transfers unless explicitly needed, and require per-transaction confirmation for every trade, swap, or crypto transfer regardless of amount.

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 (2)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The metadata and description present the skill as an autonomous financial operator that can directly manage a Coinbase portfolio without clearly requiring explicit user opt-in to autonomy or allowing the user to select a safer approval mode. Because this skill operates on live financial accounts, ambiguous autonomy can lead to unintended trades or transfers and amplifies the risk of prompt misuse, misunderstanding, or unsafe defaults.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill explicitly supports autonomous trades and on-chain transfers, but its guardrails only mention manual confirmation for trades above $100 and do not clearly warn that smaller trades and transfers can still move funds irreversibly. In a crypto context, even low-value transactions can cause real financial loss, wallet draining over repeated actions, or irreversible transfer mistakes, so the missing warning and weak confirmation model are materially unsafe.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.