ClawHub
Back to skill

Security audit

Agentic Workflow Designer

Security checks for vulnerabilities and agentic risk

Overview

This is a design-only workflow automation skill that is clearly disclosed and does not execute code, collect credentials, or persist data.

Installers should treat generated workflow specs as drafts, review them before import, and add human approval for customer-facing, financial, regulated, data-deleting, or external side-effect actions.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 28)May include surrounding context.

md
## What This Skill Does

Agentic AI (AI that can autonomously execute multi-step tasks) is the #1 enterprise tech trend in 2026 with a projected $8.5B market and 40% CAGR. Yet most teams struggle to:
- Map which workflows are actually suitable for agentic automation
- Design reliable pipelines that don't break silently
- Choose between n8n, Make, Zapier, or custom agent frameworks

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger list is extremely broad and includes generic phrases like workflow automation, process automation, and design a workflow, which can cause the skill to activate for ordinary user requests outside its intended scope. Over-broad activation increases the chance of inappropriate interception and unsolicited generation of agentic automation guidance, including for workflows involving sensitive operations.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
85% confidence
Finding

The example explicitly endorses automatic approval and rejection based on a model-derived score, which normalizes autonomous decision making with real side effects. Even though presented as an example, it can encourage users to implement threshold-based automation for outreach or gating decisions without sufficient safeguards, auditability, or fairness review.

Content

Scanner excerpt · SKILL.md (reported line 164)May include surrounding context.

[Flow]: Sequential with HITL gate [Memory]: PostgreSQL (lead history) [Error]: Retry enrichment 3x → flag for manual review [HITL]: Score > 80 auto-approves; 50-80 requires human review; <50 auto-rejects [Output]: CRM updated + email queued

text

Static analysis

No suspicious patterns detected.