Back to skill

Security audit

temporal-protect

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because scanners report an instruction to read a signing key from an environment variable and a wallet data disclosure gap.

Before installing, consider whether you are comfortable with the reported wallet data transmission and signing-key instruction. The scan covered the single reported Markdown component completely, but its findings identify meaningful concerns; the verdict is suspicious, not a claim that the author acted maliciously.

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
Vulnerability Patterns
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 workflow sends the user's wallet address to Temporal without a privacy disclosure.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The mandatory read step submits the user's wallet address to Temporal's access and account-state services, as shown by the wallet-bearing routes in the tool table. The skill warns about trading risks but does not disclose this third-party data transmission, which occurs before the later explanation and approval step.

Content

No source excerpt is available for this finding.

L48 contradicts its prohibition on private-key handling by directing the agent to read a signing key from an environment variable.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The instruction to read a signing key from an environment variable contradicts the prohibition on handling private keys and can expose a wallet credential to the agent's execution context, tool output, or logs. In a mainnet trading workflow, compromise of that credential could authorize transactions beyond the approved trade; dry-runs and human approval do not protect an exposed key.

Content

No source excerpt is available for this finding.