ClawHub
Back to skill

Security audit

Render Url

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent URL-rendering helper with a clearly optional LLM feature, though users should treat the resume-based question answering flow as sharing sensitive data with Anthropic.

Install this only if you are comfortable installing the PyPI package and Playwright browser it depends on. Use the default rendering/parsing flow for local deterministic extraction. If you enable --answer-questions, assume resume contents and extracted application-page text may be sent to Anthropic, so avoid unnecessary sensitive information and review provider privacy terms first.

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

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 75)May include surrounding context.

Always pass --detect together with --answer-questions. The LLM only ever runs on a page --detect classified APPLICATION — this keeps it from being called (and spending API credits) on a CAPTCHA, login wall, closed job, or bot-challenge page. render-and-parse --answer-questions without --detect fails immediately with an invalid_input error, before Chromium even launches.

bash
cp .env.example .env     # fill in ANTHROPIC_API_KEY
# write the candidate's resume as Markdown to resume.md (gitignored, never commit it)
render-and-parse "https://example.com/job/123" --detect --answer-questions

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The optional --answer-questions flow explicitly sends rendered page text and resume content to an external LLM provider, which can expose sensitive personal or job-application data if users are not clearly warned. In this skill context, resumes commonly contain PII and the rendered pages may contain application details, so omission of a prominent privacy/data-sharing warning creates a real confidentiality risk even though the feature is opt-in.

Content

No source excerpt is available for this finding.