Back to skill

Security audit

travel-manager

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward Chinese-language travel-planning skill, with one privacy hygiene issue around document backup advice.

Before installing, be aware that the travel guidance includes a recommendation to keep scans of passports or identity documents in cloud storage and email. If you follow that advice, use encrypted storage, strong account protection, and avoid sending unencrypted identity scans through ordinary email.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:62
Finding

Unsafe Cloud and Email Storage Guidance for Identity-Document Scans

Content
View full analysis

Vulnerability Details

File Locations:

  • SKILL.md, line 62
  • references/travel-documents.md, line 51

Vulnerability Type: Sensitive data stored without defined security controls
Risk Level: Medium

Vulnerable Snippets

SKILL.md:62:

text
Electronic backup: Store scanned copies in cloud storage and email.

references/travel-documents.md:51:

text
2. Electronic backup: Store scanned copies in cloud storage and email.

The snippets above are English renderings of the corresponding source instructions.

Technical Analysis

The Skill recommends placing scanned travel documents in cloud storage and email. In context, these documents can include passports, identity cards, visas, birth certificates, insurance records, and other personally identifiable information.

The guidance does not require:

  • Client-side or end-to-end encryption
  • Multi-factor authentication
  • Private access controls or restricted sharing
  • A trusted storage provider
  • Redaction of unnecessary document fields
  • Retention limits and secure deletion
  • Monitoring for unauthorized access
  • Separation of encryption keys from stored files

Email attachments and ordinary cloud files are frequently readable after mailbox compromise, credential theft, session-token theft, accidental public sharing, or provider-side disclosure. Creating multiple unprotected copies also expands the number of systems through which the documents may be exposed.

This is an insecure handling practice rather than programmatic exfiltration. The audited project contains no implementation that automatically uploads or transmits documents.

Attack Path

  1. A user follows the Skill's recommendation and scans a passport or another identity document.
  2. The user uploads the unencrypted scan to ordinary cloud storage and sends or saves another copy through email.
  3. The cloud account, mailbox, recovery account, authenticated ses ...[truncated 1348 chars]
Remediation
View remediation

Remediation Suggestions

Replace the current recommendation with security-specific document-handling guidance:

  1. Prefer encrypted local storage or a reputable end-to-end encrypted document vault.
  2. If cloud storage is necessary, encrypt each file locally before upload using authenticated encryption and a strong, unique passphrase.
  3. Do not send unencrypted passport or identity-document scans through ordinary email.
  4. Protect storage and email accounts with phishing-resistant multi-factor authentication and unique passwords.
  5. Disable public-link sharing and grant access only to specifically authorized accounts.
  6. Redact document fields that are unnecessary for the intended emergency or travel purpose.
  7. Store decryption keys or passphrases separately from the encrypted files.
  8. Define a short retention period and securely delete copies after travel.
  9. Review account sessions, sharing permissions, and access logs regularly.
  10. Warn users that destination authorities, airlines, hotels, and insurers may have different document requirements and that copies should only be provided through verified official channels.

Apply the same correction to references/travel-documents.md:51 so the reference document does not reintroduce the unsafe recommendation.

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (4)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest sets language: "zh-CN", which is a natural-language locale restriction. Under the policy, forcing a specific language is a violation unless the skill offers a language choice or clearly justifies the locale constraint; this file does neither.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

SQP-3 applies to all file types and covers language or locale policy violations. This markdown file presents all user-facing guidance exclusively in Chinese, with no opt-in, alternative language, or justification that the skill is specifically intended only for Chinese-speaking users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The entire checklist is written in Chinese and provides no indication that language is selectable or that the content is intentionally limited to a Chinese-speaking audience. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file contains user-facing guidance exclusively in Chinese, and there is no note that the skill is region-specific or that users may choose another language. Under the natural-language policy rule, forcing a specific language without opt-in can be a policy violation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.