ClawHub
Back to skill

Security audit

Geepers Llm

Security checks for vulnerabilities and agentic risk

Overview

This is a simple API-integration skill that clearly points requests to an external LLM service, with no local code execution or persistence.

Install only if you are comfortable sending prompts, image URLs, and text-to-speech content to api.dr.eamer.dev and possibly downstream model providers. Avoid sending secrets, credentials, regulated data, or sensitive images unless you have approved that data flow.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (6)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill explicitly routes user prompts, images, and other content to a third-party API, but the description does not warn users that their inputs leave the local environment and are processed by an external service. This creates a privacy and data-handling risk because users may unknowingly submit sensitive text, images, or generated media requests to an intermediary service that may further relay content to multiple downstream model providers.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

This endpoint sends chat messages to an external API aggregator, meaning any prompt content, attached context, and possibly sensitive business data will leave the trusted boundary. The risk is heightened by the skill's stated purpose of brokering requests across multiple providers, which increases uncertainty about downstream handling, retention, and disclosure of user data.

Content

Scanner excerpt · SKILL.md (reported line 24)May include surrounding context.

Chat Completion

text
POST https://api.dr.eamer.dev/v1/llm/chat
Headers: X-API-Key: $DREAMER_API_KEY
Body:
{

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 39)May include surrounding context.

List Models

text
GET https://api.dr.eamer.dev/v1/llm/models

Vision (image analysis)

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

The vision endpoint transmits image URLs and prompts to an external API, which can expose sensitive images, embedded text, personal data, or internal resource references outside the local trust boundary. Because image analysis often involves highly sensitive visual content, this is more dangerous than ordinary text-only requests and should be treated as significant third-party data sharing.

Content

Scanner excerpt · SKILL.md (reported line 44)May include surrounding context.

Vision (image analysis)

text
POST https://api.dr.eamer.dev/v1/llm/vision
Body: { "model": "claude-sonnet-4-5-20250929", "image_url": "...", "prompt": "Describe this image" }

External Transmission

Medium
Category
Data Exfiltration
Confidence
87% confidence
Finding

The image generation endpoint sends user prompts to a third-party API, exposing whatever text the user provides to an external processor. Although it does not inherently upload local files, prompts can still contain confidential instructions, personal data, or proprietary information that users may not realize are being shared externally.

Content

Scanner excerpt · SKILL.md (reported line 50)May include surrounding context.

Image Generation

text
POST https://api.dr.eamer.dev/v1/llm/image
Body: { "prompt": "A sunset over the ocean", "provider": "openai" }

External Transmission

Medium
Category
Data Exfiltration
Confidence
91% confidence
Finding

The text-to-speech endpoint sends raw text to an external service, which can expose sensitive content if the user includes confidential, personal, or regulated information. Since the skill provides no visible warning at the description level about third-party transmission, users may unknowingly disclose data outside their expected environment.

Content

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

Text-to-Speech

text
POST https://api.dr.eamer.dev/v1/llm/tts
Body: { "text": "Hello world", "voice": "alloy" }

Static analysis

No suspicious patterns detected.