T08 · Insecure Dependencies
Error
- Location
SKILL.md:158- Finding
Unpinned credential-handling integration gateway dependency
- Content
View full analysis
- Remediation
View remediation
Security audit
Discord Community Hub Free
Security checks for vulnerabilities and agentic risk
Overview
The skill is not proven malicious, but it needs review because a read-only Discord helper asks users to install a credential-handling gateway and includes broader command, export, and mutation language than its stated purpose needs.
Install only if you trust the `integration-gateway` source and can verify or pin the version being installed. Approve the minimum Discord OAuth scopes needed, avoid using the skill for non-Discord integrations or webhook setup, and treat any export/save/modify/reset behavior as outside the safe read-only promise unless explicitly confirmed by the platform.
Vulnerability Patterns
- Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
- Insecure DependenciesIntroduces malicious components through unsafe dependency sources
- 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
Findings (2)
T05 · Unauthorized Access and Privilege Escalation
Warning
- Location
SKILL.md:5- Finding
Read-only Skill grants broad execution and mutation capabilities
- Content
View full analysis
- Remediation
View remediation
Vulnerability Patterns
- Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- 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 (5)
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 88% confidence
- Finding
文件整体描述与用户可见说明均固定为中文，且未声明这是面向特定中文区域用户的专用技能，也未提供语言切换或用户选择。根据语言/locale 政策，若技能对外呈现时默认强制单一语言而无用户选择，可能构成自然语言层面的组织政策问题。
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
The matrix says application-command management is unsupported, yet nearby guidance discusses command-permission editing as if it were a relevant operation. Even if no direct write tool is shown here, mixing unsupported privileged actions into the skill narrative can cause an agent or operator to overestimate granted authority and invoke higher-risk integrations incorrectly.
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The skill repeatedly claims the free edition is read-only, but the capability sections explicitly advertise create/modify/reset/import/save operations via generic parameters. This contradiction can mislead an agent into attempting state-changing actions or exporting/saving sensitive data under a 'safe read-only' label, weakening trust boundaries and policy enforcement.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
The capability coverage line uses a very broad keyword list and generalized phrases that dilute the skill boundary. This makes semantic matching more permissive, increasing accidental routing of unrelated community-management or integration requests into a skill that can enumerate account, guild, and OAuth-related information.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The trigger condition is broad enough to match generic requests about API integration, interface docking, webhooks, or system connections, far beyond this skill's stated Discord read-only scope. Overbroad activation increases the chance the skill is selected in unrelated contexts, causing unnecessary credential pairing, disclosure of connected integrations, or execution of mismatched tool flows.
- Content