ClawHub
Back to skill

Security audit

Affiliate Program Search

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed affiliate-program research skill with broad activation wording but no evidence of hidden code, credential use, persistence, or destructive behavior.

Install this if you want an agent to research affiliate programs and compare commissions. Be aware it may activate on broad promotion or niche questions and will use external web/API lookups, so avoid giving private strategy details unless you are comfortable with those searches.

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)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger conditions are broad enough to activate on generic requests like finding products, niches, or 'what pays the most,' which can cause incorrect skill selection and steer the agent into affiliate-marketing behavior outside the user's true intent. In a multi-skill agent, this increases the risk of prompt-routing abuse, irrelevant external lookups, and unintended propagation of affiliate-focused outputs into downstream skills.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill enables allow_implicit_invocation without any visible trigger constraints or exclusion conditions, which can cause the agent to auto-select this skill in broader contexts than intended. Because this skill performs external research and recommendation behavior, unintended invocation could expose user queries to unnecessary external processing or cause the agent to take actions based on loose keyword matches.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

The workflow defines verdict bands as 7.5+ = "Strong Pick", 5.5-7.4 = "Worth Testing", and below 5.5 = "Skip" (L84), but the self-validation step says the top-pick verdict should match thresholds where 6+ = "Worth Considering". This is an active contradiction inside the skill documentation about how recommendations should be labeled.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.