Back to skill

Security audit

免费课程核实与分享素材 · Free Course Kit

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because scanners report that redaction verification files can persist and may temporarily contain sensitive text before all redactions are applied; the scan is partial because two referenced paths were missing.

Consider the privacy risk before installing, especially if processing certificates or other sensitive images. The scanners recommend not installing; if you proceed, check how verification files are handled and removed. Coverage is incomplete: two referenced paths were missing, although all listed components were scanned.

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
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The description promises that redaction data is discarded after local processing, but the workflow describes saved verification images and manifests without a cleanup lifecycle. This creates a misleading privacy guarantee for sensitive certificate processing; private directory permissions limit access but do not fulfill the discard promise.

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, 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.

Verification artifacts persist after completion despite the manifest's discard-after-processing promise.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The script creates verification directories with mkdtemp or an explicit path and never removes them, including after successful completion. Persisted crops and manifests contain image context, absolute paths, masked labels, and short unsalted text digests, leaving sensitive processing traces despite the stated discard promise; restrictive permissions reduce but do not eliminate this exposure.

Content

No source excerpt is available for this finding.

Target verification crops can contain requested sensitive text that has not yet been redacted, without the unredacted-verification opt-in.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

Target verification crops are saved inside the loop that applies individual redaction patches, before later sensitive regions have necessarily been masked. Because each crop includes surrounding context, an early crop can persist a nearby requested sensitive value without enabling the unredacted-verification option, undermining the privacy tool's safe default.

Content

No source excerpt is available for this finding.

The workflow includes copywriting drafts despite the manifest explicitly excluding copywriting.

Low
Category
Excessive Agency
Confidence
98% confidence
Finding

The manifest at L003 explicitly states that the skill is not responsible for copywriting (不负责文案撰写). However, L130–L135 define a copywriting-draft and delivery stage, including instructions for the generated copy; L016 also advertises copywriting drafts as an output. This expands the instructed workflow beyond the manifest's stated scope.

Content

No source excerpt is available for this finding.