Back to skill

Security audit

local-industrial-fund-analysis-writer

Security checks for vulnerabilities and agentic risk

Overview

This is mostly a writing skill, but its publishing instructions conflict with its stated no-external-write boundary.

Install only if you are comfortable with a Chinese industrial-fund writing assistant that creates local Markdown drafts. Before allowing any Tencent Docs or publishing connector use, require an explicit preview, destination confirmation, and user approval because the skill's own external-publishing boundary is inconsistent.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill's earlier safety boundary forbids external publishing/writes, but later directs publication to Tencent Docs. Contradictory instructions about external writes are dangerous because they can bypass user expectations and policy controls, leading to unauthorized disclosure of drafted content or sensitive analysis into third-party systems.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The description defines the skill as a Chinese industrial-fund analysis writing expert for a professional public-account audience, and the rest of the document consistently assumes Chinese output. There is no explicit user opt-in or alternative language choice, which can violate language/locale policy requirements when a skill forces a specific language by default.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill claims to be self-contained and not dependent on external experts or tooling, but later authorizes use of external publishing tooling. This mismatch can mislead users and downstream policy systems about the skill's actual capabilities and trust boundaries, increasing the chance of unintended external data transfer or action.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document says the skill only drafts content for user review and does not publish, yet it later includes instructions to publish via another skill or connector. This creates an authority-boundary violation: operators may believe the skill is non-actuating when it in fact contains actuating workflow guidance.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The markdown instructs the skill to write generated content as .md files in the workspace root, which affects the user's filesystem. Although the document explains output behavior, it does not include any cautionary user warning or confirmation language around automatic file creation.

Content

No source excerpt is available for this finding.