Back to skill

Security audit

GEO 中文信源地图

Security checks for vulnerabilities and agentic risk

Overview

The skill organizes article evidence, with optional publishing and signed-in browsing controlled by the user.

Core evidence export needs no accounts or external services. If you choose aimeGeo continuation, review its setup separately and confirm the final article, destination accounts and visibility before publishing; connected services may receive the content. Choose an output directory where replacing existing reports is acceptable.

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
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

The follow-through workflow instructs article distribution without warning about external disclosure or public visibility.

Medium
Category
Output Handling
Confidence
86% confidence
Finding

The copyable task instructs the agent to save the revised article in aimeGeo, generate imagery, and arrange distribution through specified accounts. This is an actionable workflow rather than a tool listing, but it does not warn that distribution may expose the article publicly; the earlier notice about external-service fees does not address that consequence.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read, file_write, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/self_test.py (reported line 34)May include surrounding context.

python
return json.loads(files["校验汇总.json"])

    def run_cli(self, input_path, output_path, cwd):
        return subprocess.run(
            [sys.executable, "-B", str(ROOT / "scripts" / "evidence_table.py"),
             "--input", str(input_path), "--out-dir", str(output_path)],
            cwd=cwd, capture_output=True, text=True, encoding="utf-8", check=False,