ClawHub
Back to skill

Security audit

Sharpagent Intelligence Monitor

Security checks for vulnerabilities and agentic risk

Overview

This skill is a public-source intelligence briefing workflow with disclosed web collection, scoring, verification, and local archiving behavior.

Before installing, be aware that the skill gathers and summarizes public web sources and saves dated briefing files locally. Treat quick-scan items as unverified leads, and periodically delete archived briefings if the monitored topics are sensitive.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

The workflow allows scan-tier candidates to 'skip verification,' meaning unverified information can still be included in outputs and influence downstream decisions or trend analysis. In an intelligence-monitoring context, this increases the risk of propagating misinformation, marketing manipulation, or adversarial content from noisy sources, especially because the system is designed to autonomously collect and summarize external material.

Content

Scanner excerpt · SKILL.md (reported line 263)May include surrounding context.

md
5. 🔄 Can we verify it independently?

Watch-tier candidates get a lightweight check (source + logic).
Scan-tier candidates skip verification.

### Step 4: Compute Final Confidence

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill explicitly persists collected intelligence to a dated file path, but provides no user-facing warning, consent step, or retention guidance. Even if the data is mostly public-source content, archived briefings can accumulate sensitive research interests, monitored entities, URLs, and derived analysis that may create privacy, confidentiality, or governance issues over time.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.