Back to skill

Security audit

Course Paper Writer · 课程论文写作

Security checks for vulnerabilities and agentic risk

Overview

ClawScan’s verdict is benign with medium confidence: the scan found no AIG issues, but Skillspector reported purpose and scope mismatches, and its review was partial because three references were unresolved.

The verdict is benign, not risk-free. Before installing, consider whether the package-audit and credential-scanning behavior fits your needs, and note the undeclared tool scope and possible citation-checking limitation. Skillspector inspected all 18 listed components, but its analysis was marked partial because three references in `SKILL.md` did not resolve, so the evidence is 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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied script has a materially different primary purpose from the declared academic-writing workflow. It audits a skill directory for publication readiness rather than reading assignment documents, researching literature, writing papers, or generating formatted .docx files. Its recursive credential and packaging scans are unrelated to the declared writing tasks, not merely implementation details of those tasks. The script performs local reads and console reporting; no network transmission, file modification, or execution of detected tools is shown. This assessment applies to the supplied chunk, not any unseen components of the skill.

Content

No source excerpt is available for this finding.

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

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.

Recursively scanning files for credentials is unrelated to the declared academic-writing workflow.

Medium
Category
Excessive Agency
Confidence
90% confidence
Finding

The manifest describes reading assignment requirements, researching sources, and generating formatted papers. L207–L222 instead recursively read files under the selected directory, search for credentials, and report a matched credential prefix; this publishing-audit capability is documented by the script but is not justified by the skill's stated purpose.

Content

No source excerpt is available for this finding.

The helper claims to check numbering continuity but reports consecutive numbering without inspecting reference numbers (L008, L092–L093).

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The continuity check tests whether reference bodies are nonempty after removing numeric prefixes, rather than comparing the reference numbers. Duplicate, missing, or out-of-order numbers can consequently receive a false continuity assurance, undermining a validation step central to this skill's promised document integrity.

Content

No source excerpt is available for this finding.