ClawHub
Back to skill

Security audit

Chainflip Swap

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed crypto-swap skill, but it relies on a mutable remote service and optional remote prompt for irreversible real-funds transfers, so it belongs in Review rather than automatic install.

Install only if you are comfortable trusting the Chainflip BaaS MCP endpoint for deposit-address generation and swap status. Before funding any address, independently verify the endpoint, assets, recipient address, refund address, minimum price, fees, and deposit address, and avoid loading remote prompt guidance or sending an optional partner API key unless you intentionally need it.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (2)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:121
Finding

Remote MCP Prompt Can Inject Mutable Instructions into the Agent Session

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Error
Location
SKILL.md:38
Finding

Unpinned Remote MCP Service Controls Irreversible Cryptocurrency Deposit Addresses

Content
View full analysis
Remediation
View remediation

SkillSpector

Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger guidance is broad enough to fire on ambiguous phrases like 'what's the rate' or generic status questions, which can invoke a real-funds crypto swap skill when the user may only want information. In this context, unintended invocation is more dangerous than usual because later tool calls can initiate irreversible financial workflows and expose deposit addresses for real asset transfers.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% 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 96)May include surrounding context.

md
### Step 4, send funds
Instruct the user to send the source asset to the returned deposit address. This is a plain
native send, no approval or contract interaction. If the swap cannot meet the minimum price,
funds are returned to `refundAddress`.

### Step 5, monitor

Static analysis

No suspicious patterns detected.