SS需求→Teambition任务

• Outcome: Review
• Latest audit: Jul 9, 2026 · 1:15 PM UTC
• Version: 1.0.4

Overview

The skill matches its automation purpose, but it needs review because it ships a real-looking SaleSmartly API key and moves raw customer support conversations into local files and Teambition tasks.

Install only if you administer the relevant SaleSmartly and Teambition workspaces. Remove and rotate the packaged SaleSmartly API key, replace scripts/config.json with your own private config, protect the Teambition token, review generated tasks before enabling cron, and confirm your customer-data policy allows these conversations to be stored locally and copied into Teambition.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• 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)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 95% confidence
• Finding: The skill clearly instructs the agent to read and write local files and access external services, but it does not declare those permissions explicitly. Undeclared capabilities weaken user and platform transparency, making it harder to assess data access and exfiltration risk before execution, especially since the workflow handles customer conversation data and external task creation.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 85% confidence
• Finding: The declared purpose describes conversation collection, AI analysis, and task creation, but it omits material operational behavior: storing raw session data locally for 7 days and persisting incremental state in state.json. This mismatch can mislead users about data retention and local persistence, which is security-relevant because customer support conversations may contain sensitive or regulated information.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill description does not warn that customer conversation content will be analyzed by AI and used to create tasks in an external system. This is dangerous because support chats often include personal, confidential, or commercially sensitive data, and forwarding or summarizing them into Teambition can create unauthorized disclosure, privacy, and compliance issues.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The guide directs the agent to copy customer conversation content, customer identifiers, and real agent names into Teambition tasks, but it does not require user consent, data minimization, masking, or any warning that this information will be sent to an external system. Because the skill’s purpose is to transform support chats into project-management artifacts, this context makes the issue more dangerous: sensitive support data is likely to include personal, business, or account details, and the workflow operationalizes bulk transfer of that data.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The guide instructs users to retrieve a Teambition User Token and copy it directly, but provides no warning that this token is a bearer credential that grants API access as the user. In a skill whose purpose is to automate task creation through MCP, this omission increases the chance that operators mishandle, share, or log the token, enabling unauthorized access if it is exposed.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The configuration example embeds the Teambition User Token directly in a plaintext local config file, creating persistent storage of a high-value credential without any warning about filesystem access, backups, or accidental disclosure. Because this skill enables API actions such as creating and modifying tasks, compromise of the stored token could let an attacker access project data and perform actions as the victim user.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The script writes customer names, chat_user_id values, project identifiers, and full chat transcripts to local JSON files in cleartext. In this skill context, the data consists of customer support conversations intended for later AI processing, so storing it on disk without access controls, minimization, or retention safeguards increases privacy and data-leak risk if the host, workspace, backups, or logs are accessed by unauthorized parties.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.