Back to skill

Security audit

geopolitics-intel

Security checks for vulnerabilities and agentic risk

Overview

The report service fits the skill’s purpose, but broad activation rules and contradictory privacy statements could cause unexpected sharing of user-derived topics.

Use this skill only for explicit geopolitical report requests, and review the sanitized topic before submission. The external service receives a topic derived from your request despite the privacy summary’s claim; avoid confidential subjects or business details.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 9)May include surrounding context.

md
openclaw:
    requires:
      bins:
        - curl
    envVars:
      - name: GEOPOLITICS_TOKEN
        required: false

Everyday words are defined as standalone activation triggers without requiring geopolitical context.

Medium
Category
Trigger Abuse
Confidence
98% confidence
Finding

The activation rule treats ordinary words such as 'impact' and 'will it' as sufficient triggers without requiring geopolitical context. Combined with the instruction to immediately create an external report, this can redirect unrelated requests and transmit user-derived topics unnecessarily; the narrower skill description and redaction requirements reduce but do not eliminate this risk.

Content

No source excerpt is available for this finding.

The data-handling disclosure claims an exclusive upload list that omits the user-derived topic actually sent to create_report.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

The data-handling section claims an exhaustive upload list and says the skill does not collect user input, but the default workflow sends a user-derived topic to an external report service. Although earlier sections disclose topic submission and require redaction, the contradictory privacy summary can mislead users; a sanitized topic can still reveal sensitive interests or business concerns, and server-side masking occurs only after transmission.

Content

No source excerpt is available for this finding.