ClawHub
Back to skill

Security audit

Roadrunner

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Beeper Desktop CLI helper for reading and acting on chats, with sensible safety instructions around private data and message-sending.

Install this only if you want an agent to help operate Beeper Desktop. Treat it as sensitive because it can read private chats and contacts and, when explicitly asked, send messages, upload files, create chats, archive chats, and set reminders. Prefer the documented read-only and dry-run modes for exploratory work, and keep auth tokens configured locally rather than pasting them into chat.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • 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)

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 100)May include surrounding context.

md
- Message edit: `rr messages edit "!chatid:beeper.com" "<message-id>" "Updated text"`
- Message react/unreact: `rr messages react "!chatid:beeper.com" "<message-id>" "👍"` / `rr messages unreact "!chatid:beeper.com" "<message-id>" "👍"`
- Upload + send file: `rr messages send-file "!chatid:beeper.com" ./photo.jpg "See attached"`
- Create chat: `rr chats create "<account-id>" --participant "<user-id>"`
- Start chat from merged contact hints: `rr chats start "<account-id>" --email "alice@example.com" --full-name "Alice"`
- Archive/unarchive: `rr chats archive "!chatid:beeper.com"` / `rr chats archive "!chatid:beeper.com" --unarchive`
- Reminder mutations: `rr reminders set "!chatid:beeper.com" "2h"` / `rr reminders clear "!chatid:beeper.com"`

Static analysis

No suspicious patterns detected.