ClawHub
Back to skill

Security audit

sf-scrapper

Security checks for vulnerabilities and agentic risk

Overview

This skill is not malicious, but it uses a live authenticated SuccessFactors browser session to collect sensitive employee HR data with limited safeguards.

Install only if you are authorized to view the requested employee records in SuccessFactors. Treat returned names, emails, managers, locations, and phone numbers as sensitive HR data; ask for only the fields needed, avoid broad batch use unless approved, and do not rely on this skill as an auditable or least-privilege HR integration.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
- **Never use OData API, REST endpoints, or any programmatic API.** This skill is purely browser-based scraping.
- Always use `profile="chrome"` — never `profile="openclaw"` (we need the user's authenticated session).
- Be patient with page loads — SF can be slow. Use snapshots to verify page state before extracting.
- Respect the user's session — don't navigate away from SF without warning.
Confidence
80% confidence
Finding
Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs scraping employee details like email, manager, location, and phone from an authenticated SuccessFactors session but does not present an explicit privacy and authorization warning up front. Because the workflow targets HR data in a live user session, the lack of consent, least-privilege, and data-handling guidance materially increases the chance of unauthorized access, over-collection, or disclosure of sensitive employee information.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger phrases are broad enough that the skill may activate on loosely related requests involving employee IDs or SuccessFactors, causing browser automation against a live authenticated HR session without sufficiently explicit user intent. In this context, over-triggering is more dangerous because the skill accesses sensitive employee data from a logged-in enterprise system, increasing the risk of unnecessary disclosure or collection of personal information.

Static analysis

No suspicious patterns detected.