ClawHub
Back to skill

Security audit

TradingCalc

Security checks for vulnerabilities and agentic risk

Overview

The skill provides a clearly disclosed trading-calculator integration, with privacy and calculation assumptions users should consider.

Limit use to supported trading calculations and share only necessary inputs you are comfortable sending to TradingCalc. Exclude passwords, private keys, and identifying account details. Confirm fee and margin assumptions before relying on results; formula tests and signatures do not validate your particular trading setup. The remote implementation and MCP-specific retention were not audited.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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)

The activation instruction extends beyond trading to any actionable numerical question.

Medium
Category
Trigger Abuse
Confidence
98% confidence
Finding

The instruction to use this skill whenever a question needs an actionable number extends beyond its trading-specific purpose. An agent following that instruction could unnecessarily invoke external tools for unrelated calculations, potentially disclosing inputs or applying inappropriate financial models; the trading-specific examples constrain this risk but do not explicitly override the broad activation rule.

Content

No source excerpt is available for this finding.

The workflow sends financial inputs to an external service without a privacy warning.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The workflow directs the agent to call an external MCP service without requiring a privacy check or ensuring the user understands that financial inputs leave the agent's environment. Position details, portfolio return series, and wallet addresses can reveal sensitive financial information, although routine public or hypothetical calculator inputs carry substantially less risk. Response signatures and calculator verification do not address confidentiality or the service's handling of submitted data.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.