ClawHub
Back to skill

Security audit

institutional-13f-tracker

Security checks for vulnerabilities and agentic risk

Overview

This is a read-only financial research skill with no local file access, persistence, or trading authority, though users should treat its cross-source “convergence” language as research context rather than advice.

Install only if you are comfortable sending ticker and manager research queries to SentiSense with a SentiSense API key. Treat the output as lagged 13F research data, not as trading advice, and avoid relying on the skill's convergence language as a buy, sell, or confidence signal.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The phrase encouraging users to treat cross-skill convergence as a 'high-conviction read' pushes the skill beyond neutral data retrieval into signal-generation language that can steer trading decisions. In a financial research skill, this weakens the stated 'informational only' boundary and can cause downstream agents to present composite investment suggestions as if they were sanctioned by the skill.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The workflow instructs the agent to surface converging institutional, congressional, and insider signals, which operationally resembles multi-factor trade idea generation. Even without direct order execution, this creates a pathway for an agent to provide actionable investment-style guidance while the document simultaneously claims it is not advisory.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The 'When to Use' section presents general natural-language prompts such as 'Who owns $NVDA?' and 'What did Berkshire Hathaway buy and sell last quarter?' but does not define any explicit activation boundaries or exclusion conditions. In a manifest/markdown skill description, this can make invocation scope overly broad for ordinary stock-research questions that may or may not require this specific skill.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.