ClawHub
Back to skill

Security audit

Blave Quant

Security checks for vulnerabilities and agentic risk

Overview

The trading features have confirmation safeguards, but marketplace strategies run downloaded code without adequate protection for local files and credentials.

A.I.G

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
references/marketplace.md
Finding

Third-party strategy code runs with the agent's local privileges without isolation

Content
View full analysis

Impact

A malicious strategy could read locally accessible .env credentials, modify files, or issue authenticated exchange requests outside the agent's conversational confirmation flow. These are consequences of unrestricted execution, not observed behavior of a downloaded payload.

Static analysis

No suspicious patterns detected.