Back to skill

Security audit

HyperGrok Desk Post Trade Review

Security checks for vulnerabilities and agentic risk

Overview

ClawScan labels this skill benign because the completed checks found no major security indicators and its two reported issues concern decision autonomy and a broad trigger; confidence is medium because Skillspector coverage was partial.

The verdict is benign, but it is not a guarantee of safety. Consider whether you are comfortable with the approval wording and broad activation phrase. Skillspector scanned all three listed components, but its analysis was marked partial because several local references could not be resolved; treat the evidence as incomplete.

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
  • 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)

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 66)May include surrounding context.

md
Grade:

- **Process:** clean / minor break / major break, with the specific stage named. A major break is any send without PASS or without approval by id or a passing mandate check, any missing protection, any resend on unknown result, or any limit breached.
- **Outcome:** result in R and USD, stated without adjectives.

Then **one thing**: a leak, a control that worked, or a break, chosen because it is repeatable. Not a list.

The activation phrase "how did that go" can trigger a trade review for unrelated everyday requests.

Low
Category
Trigger Abuse
Confidence
92% confidence
Finding

The description explicitly activates the skill when the user asks "how did that go," without requiring a reference to a trade or desk activity. Although the surrounding description concerns trading, this invocation condition can also match questions about unrelated tasks.

Content

No source excerpt is available for this finding.