ClawHub
Back to skill

Security audit

Google Reviews

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent Google review research and monitoring guide with disclosed local storage and Google API/page access, not hidden or destructive behavior.

Before installing, expect this skill to create ~/google-reviews/ if you enable monitoring and to store brand watchlists, snapshots, reports, cadence, and alert preferences there. Use the stay-quiet activation option if you only want explicit invocation, and avoid placing credentials or private customer identifiers in the memory files.

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)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The activation scope is broad enough to match many ordinary requests about reviews, reputation, or monitoring, which can cause the skill to invoke when the user did not clearly request it. Over-broad activation increases the chance of unintended data capture, persistence, or behavior changes, especially because the skill later instructs storing reusable context to local memory.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The future-session activation language allows self-invocation under vague conditions like reputation or heartbeat/reporting-related requests, without a strict user-consent boundary. This can make the skill persistently re-engage across sessions and influence behavior unexpectedly, which is more concerning here because the skill also stores activation preferences and monitoring context in local files.

Unbounded Resource Access

Medium
Category
Excessive Agency
Content
1. Define owner and permissions before first fetch for each source.
2. Save connector status per brand: `active`, `degraded`, `blocked`.
3. Keep last-success timestamp and last-error reason for troubleshooting.
4. Use source-specific retry policy; do not retry indefinitely.

## Refresh Windows
Confidence
75% confidence
Finding
Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.

Static analysis

No suspicious patterns detected.