Back to skill

Security audit

时光梳理·每日行程梳理

Security checks for vulnerabilities and agentic risk

Overview

Review is recommended because the skill can replace evidence-backed locations with inferred ones despite promising to preserve that evidence.

Work on copies in a private folder and back up the baseline workbook before every run. Leave transit-city pruning disabled and independently check inferred days against original evidence before using the timeline for reimbursement or residency support. Protect GPS exports and delete unnecessary intermediates.

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
  • 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 (3)

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

The GPS-extraction workflow creates a sensitive location-history file without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The recursive extraction creates a persistent plaintext collection of precise GPS coordinates, timestamps, and potentially identifying source paths without specifying access restrictions or retention safeguards. Location processing is authorized by the skill's purpose, but aggregating years of records magnifies exposure if the working directory is shared or automatically synchronized; no external transmission is shown.

Content

No source excerpt is available for this finding.

Transit-city pruning discards verified evidence days despite documentation promising to retain them.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

When more than two days in a transit-city segment have supporting evidence, the code replaces the evidence-day set with the segment's first and last indices, then rewrites other days to HOME_CITY and replaces their city lists. This optional configuration can therefore discard verified intermediate locations and produce misleading records, particularly consequential when outputs are used for residency, tax, or reimbursement support.

Content

No source excerpt is available for this finding.