捷帮数据格式化

• Outcome: Review
• Latest audit: Jun 19, 2026 · 2:33 AM UTC
• Version: 1.0.0

Overview

The skill works as a formatter, but it silently sends all user-provided data to an external service using an admin-style credential.

Review before installing. Do not use this skill with secrets, production SQL, internal configuration files, customer data, proprietary XML/HTML, or other sensitive content unless you are comfortable sending that data to the jiebang.site service and understand the credential and retention practices.

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Description-Behavior Mismatch

Medium

• Confidence: 98% confidence
• Finding: The skill is presented as a local data-format conversion utility, but every operation sends the user-provided content to an external domain. This creates an unexpected data exfiltration path for potentially sensitive JSON, YAML, XML, SQL, HTML, or cron content, especially because users would reasonably expect these transformations to occur locally.

Context-Inappropriate Capability

Medium

• Confidence: 96% confidence
• Finding: The code retrieves an admin credential from the environment for a simple formatting skill, which is a privilege mismatch for the stated functionality. If the external service or this skill is abused, the credential may authorize broader backend capabilities than necessary, increasing blast radius beyond mere formatting operations.

Context-Inappropriate Capability

Medium

• Confidence: 95% confidence
• Finding: The conversion and formatting behavior depends on outbound requests to a third-party service even though the declared purpose does not require network access. This unnecessarily exposes user data to an external system and makes confidentiality depend on an unrelated remote service's security and handling practices.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: User-supplied content and the authentication header are sent to an external API with no user-facing disclosure or warning. Because formatter inputs often contain secrets, configs, SQL queries, or internal documents, silent transmission can leak sensitive material outside the expected trust boundary.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: This XML operation transmits arbitrary user XML to the remote service without explicit notice. XML commonly contains structured business data or configuration files, so silent export to a third party can violate user expectations and data governance requirements.

Missing User Warnings

Medium

• Confidence: 99% confidence
• Finding: SQL text is sent to an external API without disclosure, which is especially risky because SQL often embeds table names, schema details, tenant identifiers, literals, or even credentials. This can expose internal database structure and sensitive business logic to an external service unexpectedly.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The remaining formatter/converter endpoints also forward user inputs externally without any disclosure. Even seemingly low-risk data like HTML snippets, encoded values, or cron expressions may contain proprietary logic, tokens, or internal operational details, so the silent network transfer remains a real confidentiality issue.

Static analysis

No suspicious patterns detected.