ClawHub
Back to skill

Security audit

redfin-fpx

Security checks for vulnerabilities and agentic risk

Overview

The Redfin integration fits its stated purpose, but its saved-homes example can leave private account data in an inadequately protected temporary file.

Review the temporary-file handling before using saved homes: pipe responses directly into extraction commands, or use private, uniquely created files with cleanup. Approve only Redfin fetch access and revoke pairing when finished. Both bundled documents were reviewed; external CLI and extension binaries remain outside this assessment.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

The saved-homes workflow writes authenticated favorites-page HTML to a fixed temporary file without warning about local persistence or overwriting.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

The workflow saves authenticated favorites-page HTML to the predictable path /tmp/favs.html without restricting permissions or removing it afterward, potentially exposing private saved-home information to other local users under permissive defaults. On shared systems, an attacker could also pre-create a symlink at that path and cause shell redirection to overwrite a file writable by the executing user, subject to operating-system protections. Authentication makes this more sensitive than the skill's public listing examples, although there is no evidence of deliberate exfiltration.

Content

No source excerpt is available for this finding.