ClawHub
Back to skill

Security audit

Pudding Data Story

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent data-storytelling workflow with a small local audit script; the main caution is broad automatic activation, not hidden or dangerous behavior.

Install this if you want strong guidance for interactive data journalism and scrollytelling. Be aware that it may be selected automatically for broad data-visualization or Chinese-language data-story requests, so explicitly name another workflow when you do not want this editorial framing.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description contains very broad trigger phrases such as general requests for websites, audits, scrollytelling, interactive visualization, and 'don't make it look like a normal chart,' which can cause the skill to activate for many adjacent tasks. Over-broad auto-invocation can override more appropriate specialized skills or inject irrelevant instructions into unrelated user workflows, reducing control and increasing the chance of unsafe context blending.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill explicitly claims Chinese-language triggering for broad request categories without indicating user opt-in or a narrow routing boundary. This can cause the skill to be selected simply because a user writes in Chinese about data visualization, even when they did not request this specific workflow, creating unwanted instruction injection and misrouting across languages.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill enables implicit invocation (allow_implicit_invocation: true) without any visible narrowing conditions, which can cause the agent to auto-select this skill in contexts the user did not clearly intend. Because this skill influences editorial framing, data storytelling, implementation architecture, and audits, unintended activation could steer outputs, expose broader tool behavior, or create prompt-routing confusion across loosely related requests.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.