sales-intelligence-outreach

• Outcome: Review
• Latest audit: May 29, 2026 at 9:35 PM UTC
• Version:
  1.0.0

Overview

This sales outreach skill is transparent about researching prospects and posting content, but it lacks clear approval, account-scope, and data-handling controls before real external publishing.

Install only if you are comfortable enforcing your own review process before any post or message is sent. Use limited-scope platform accounts, avoid uploading unnecessary personal or confidential prospect data, verify the external skills it depends on, and treat all generated outreach as drafts until a human approves the exact content and destination.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (7)

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The README explicitly promotes one-click publishing to multiple social platforms and presents '已发布的小红书/公众号内容' as an output, but only gives a brief generic note to review content before publishing. In an agent skill, that framing can normalize autonomous posting or outreach on a user's behalf without clear consent, approval gates, account-scope limits, or anti-spam safeguards, increasing the risk of unauthorized actions, policy violations, and reputational damage.

Missing User Warnings

Low

• Confidence: 86% confidence
• Finding: The workflow describes collecting company updates, financing information, and generating personalized outreach based on prospect analysis, but the privacy warning is minimal and does not explain acceptable data sources, handling rules, retention, or restrictions on personal/contact data. In a sales-intelligence context, this can lead users to over-collect or process prospect information in ways that violate privacy expectations, internal policy, or applicable law.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The trigger list contains broad sales terms such as '获客', 'B2B销售', and '客户触达', which are common across many normal user conversations. This increases the chance of unintended invocation, causing the skill to activate in contexts where the user did not explicitly request automated prospecting or outreach behavior.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill prominently describes end-to-end automation that includes publishing to Xiaohongshu, WeChat, LinkedIn, and Weibo, but does not provide a clear user-facing warning that content may be posted or messages sent on the user's behalf. In a sales outreach context, accidental or implicit execution is especially risky because it can result in unauthorized external communications, spam-like behavior, reputational harm, or policy violations across multiple platforms.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The trigger list uses very broad sales and lead-generation phrases, which can cause the skill to activate for ordinary business requests without clearly signaling that it may perform customer intelligence gathering and outreach actions. In this skill's context, that ambiguity is more dangerous because the workflow later leads to automated content generation and external publishing, increasing the chance of unintended activation for sensitive or high-impact actions.

Missing User Warnings

High

• Confidence: 96% confidence
• Finding: The top-level description promotes one-click multi-platform publishing and customer outreach but gives no warning that the skill can post to external accounts or create public-facing content. That is dangerous because users may invoke the skill expecting analysis assistance, while the workflow is positioned to perform consequential actions affecting third-party platforms, brand reputation, and potentially privacy-regulated customer data.

Missing User Warnings

High

• Confidence: 98% confidence
• Finding: The workflow includes a direct 'one-click' outreach stage that publishes content to Xiaohongshu, WeChat official accounts, and LinkedIn without mentioning review, consent, privacy checks, or safeguards around account actions. In this context, the risk is elevated because earlier stages gather customer intelligence and generate personalized materials, so the final stage can operationalize potentially incorrect, sensitive, or noncompliant content at scale.

VirusTotal

63/63 vendors flagged this skill as clean.