100Critical
This skill allows remote code execution,
Behavior Analysis
Description verified — behavior matches claim
Claims to do
red-crawler-ops: Use this skill when you need to operate the `red-crawler` CLI from an OpenClaw workflow. It is the portable wrapper for the repo's existing crawler runtime, not a separate crawler implementation.
Actually does
This skill executes various `red-crawler` CLI commands via `uv run red-crawler` to interact with the Xiaohongshu platform. It performs actions like crawling user profiles, logging in, collecting data, and generating reports. It accesses local SQLite databases and Playwright storage state files, and can install necessary tools like `uv` and `git` by contacting their respective download URLs.
Skill Info
Namebatxent/rednote-contacts
Registryclawhub
Version1.0.0
PURLpkg:clawhub/batxent/rednote-contacts@1.0.0
Downloads34
Source
SHA-25608b2b72ef34f...
SHA-1476e503cd4b1...
MD518b222388ad6...
TLSHc122c517128c...
ssdeep192:PutnGoJz...
Size10,410 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
ClawHubDocs
openclaw skills install batxent/rednote-contactsAssessments (1)
AI Agent Traps ↗Remote code download and execution detectedcriticalcommand execution
Remote code download and execution detected
100% confidenceline 206
curl -LsSf https://astral.sh/uv/install.sh | sh
Arbitrary Command Execution with Path Manipulationcriticalcommand execution
The skill executes `uv run red-crawler` commands with numerous parameters accepting file paths. An attacker could manipulate these paths to read/write arbitrary files, overwrite system files, or exfiltrate data. Setup instructions also include `sudo apt-get install` and `curl | sh`.
100% confidence
`uv run red-crawler crawl-seed --db-path "./data/red_crawler.db" --output-dir "./output"`, `sudo apt-get install -y git python3 python3-pip`, `curl -LsSf https://astral.sh/uv/install.sh | sh`
Handling of Sensitive Authentication Datahighcredential theft
The `login` command saves Playwright storage state (`state.json`) containing authentication tokens. If an attacker can control the `--save-state` path or access this file, it could lead to session hijacking or credential theft.
100% confidence
`uv run red-crawler login --save-state "./state.json"`, "Path to save storage state"
Data Exfiltration of Sensitive Informationhighdata exfiltration
The skill's primary function is to crawl and extract sensitive 'contact information (emails, etc.)' and generate reports. If an attacker can control the output directories, this sensitive data could be directed to an attacker-controlled location or made accessible for exfiltration.
100% confidence
"Outputs: accounts.csv, contact_leads.csv", "Extracted contact information (emails, etc.)", `--output-dir`, `--report-dir`
Potential for Resource Abuse via Crawling Parametersmediumresource abuse
Crawling commands (`crawl-seed`, `collect-nightly`) have parameters like `max-accounts`, `max-depth`, and `crawl-budget`. An agent could be instructed to set these to excessively high values, leading to denial-of-service or excessive resource consumption on the host system or target website.
90% confidence
`--max-accounts 5`, `--max-depth 2`, `--crawl-budget 30`
Privilege Escalation via Setup Commandsmediumprivilege escalation
The environment setup instructions explicitly include `sudo apt-get install` commands. If the agent is permitted to execute these, it could be used to install malicious packages or modify system configurations with elevated privileges.
90% confidence
`sudo apt-get update`, `sudo apt-get install -y git python3 python3-pip`
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/clawhub/batxent/rednote-contacts)HTML
<a href="https://mondoo.com/ai-agent-security/skills/clawhub/batxent/rednote-contacts"><img src="https://mondoo.com/ai-agent-security/api/badge/clawhub/batxent/rednote-contacts.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/clawhub/batxent/rednote-contacts.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow