Back to skill

Security audit

Financial Private Equity

Security checks for vulnerabilities and agentic risk

Overview

This private-equity workflow is coherent overall, but it should be reviewed because it can direct an agent to search Gmail and Slack without an explicit consent or scope limit.

Install only if users understand that deal sourcing may inspect connected Gmail and Slack. Require explicit confirmation before using those sources, limit searches to named companies or founders, prefer metadata or existence checks, and avoid quoting or summarizing message contents unless the user asks for it.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
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)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

In a markdown skill description, trigger phrases should be specific enough to avoid accidental invocation. "find companies" is generic and could match many unrelated requests, especially outside the intended private-equity sourcing context.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The deal-sourcing workflow instructs the agent to search Gmail and Slack for prior correspondence, which expands access into sensitive communications beyond the core task of identifying companies and drafting outreach. Even if relationship-checking is business-relevant, the skill does not constrain scope, require explicit consent, or limit what data may be retrieved and summarized, creating a real privacy and data-exposure risk.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill tells the agent to inspect email and Slack for prior discussions without any privacy warning, consent gate, or least-privilege guidance. In a PE workflow, those systems may contain confidential deal discussions, personal data, and privileged material, so silent querying or summarization could expose sensitive information to the wrong user or context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

Natural-language policy requires avoiding forced language or locale constraints without user opt-in or clear justification. The document switches into Chinese for section headings and descriptive text with no indication that the user can choose language or that the skill is region-specific.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

This phrase is common conversational language and does not clearly indicate PE returns modeling on its own. Without stronger constraints, it may cause unintended activation for unrelated rough calculations.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.