ClawHub
Back to skill

Security audit

Bulgaria

Security checks for vulnerabilities and agentic risk

Overview

This is a Bulgaria travel-planning skill that gives destination advice and keeps local trip notes; the main caution is reviewing what personal itinerary details you allow it to remember.

Install only if you are comfortable with the agent keeping Bulgaria trip notes in ~/bulgaria/ and an activation preference in memory. Avoid saving exact travel dates, bookings, mobility constraints, or budget details unless they are useful, and choose the non-proactive activation option if you only want help when you ask.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (7)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill explicitly instructs creating a persistent file containing trip context, preferences, food notes, mobility or hiking notes, and booking information, but provides no notice, consent flow, retention limit, or guidance on handling potentially sensitive personal data. Even though the data is travel-oriented, it can still reveal location history, schedules, health-related mobility constraints, and behavioral preferences, creating a privacy risk if stored silently or reused unexpectedly.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation prompt is overly broad and encourages persistent future intervention whenever 'Bulgaria comes up,' which can cause the skill to activate in contexts the user did not specifically intend. This increases the chance of unwanted behavior and makes consent to future use ambiguous, especially when paired with memory storage instructions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill instructs saving activation preferences to main memory without any user-facing notice about retention, scope, or how that memory will be used later. Storing behavioral preferences without transparent consent creates a privacy risk and can lead to persistent personalization the user did not fully understand or intend.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill directs internal capture of potentially sensitive travel details such as dates, booked places, regions of interest, budget, and hard constraints without warning the user that this personal information may be retained. Travel data can reveal habits, future absence, spending profile, and personal circumstances, so silent storage creates meaningful privacy and profiling risks.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · accommodation.md (reported line 56)May include surrounding context.

md
- Choosing Golden Sands by default when the user wants Varna city life
- Expecting the beach zone to feel local in peak season
- Staying only in resort compounds and missing the city entirely
- Visiting in shoulder season without checking what is actually open

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · varna.md (reported line 49)May include surrounding context.

md
- Choosing Golden Sands by default when the user wants Varna city life
- Expecting the beach zone to feel local in peak season
- Staying only in resort compounds and missing the city entirely
- Visiting in shoulder season without checking what is actually open

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The heading explicitly scopes the content to Bulgaria, which can be a locale-specific constraint. Because the file does not indicate that this regional focus is optional or selected by the user, it may conflict with the policy against forcing a specific locale without opt-in.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.