ClawHub
Back to skill

Security audit

捷帮图片工具

Security checks for vulnerabilities and agentic risk

Overview

The skill matches its image/QR purpose, but it silently sends user content to an external service using an admin-key credential without clear user-facing disclosure.

Review before installing. Use this only for non-sensitive images and QR contents you are comfortable sending to JieBang's service. Avoid private photos, internal or signed URLs, credentials, tokens, and confidential QR payloads. Prefer a version that clearly discloses remote processing and uses a narrowly scoped API key instead of an admin-key credential.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill retrieves an admin-scoped credential from an environment variable and sends it on every request to an external service, even though the advertised functionality is only image conversion and QR generation. Using an administrator key for routine user-driven operations violates least-privilege and increases the blast radius if the remote endpoint is abused, logs headers, or the skill is repurposed for unintended API actions.

Vague Triggers

Medium
Confidence
72% confidence
Finding
The trigger phrases are broad enough that the skill may be invoked for loosely related image or QR tasks without clear boundaries on what data will be sent through the tool. Over-broad routing increases the chance of accidental invocation on sensitive user content, especially when the tool accepts URLs and embedded image data.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill accepts user-provided image URLs and Base64 image data but does not warn that such content may be transmitted to or processed by the tool. This omission can lead to inadvertent exposure of private images, internal URLs, signed links, or embedded metadata, and remote URL handling can also increase SSRF-like risk depending on implementation.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The image conversion function transmits user-supplied image content or image URLs to an external domain without any disclosure, consent flow, or local-processing alternative. Images may contain sensitive visual content or embedded metadata, so silent third-party transfer creates a privacy and data-handling risk beyond what a user may reasonably expect.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The QR generation function sends user-provided text to an external service via GET and, on failure, POST, without warning the user. QR payloads often contain sensitive links, contact data, tokens, or internal identifiers, and using GET additionally risks exposure in URLs, logs, proxies, browser history, and intermediary infrastructure.

Static analysis

No suspicious patterns detected.