Encrypted File Reader

• Analyzed: May 11, 2026, 4:57 AM
• Findings: 1
• Version: 1.0.0
• Hash: 33cf5129161a2ca5908c1b4d8efd915b4ebda50c8b956ed0c83c04225c978ba8
• Source: palm
• External report: View on VirusTotal

Findings (1)

The skill is classified as **suspicious**. The `read_file.py` script, while ostensibly designed for reading various file types, includes a very extensive list of `TEXT_EXTENSIONS` that encompasses highly sensitive files such as `.env`, `.pem`, `.key`, `.crt`, `.ssh`, `.aws`, `.kubeconfig`, `.npmrc`, `.gitconfig`, `.bashrc`, `.history`, and many other configuration, credential, and log files. The script's primary function is to read the content of any specified local file and print it to standard output. While the skill's documentation (`SKILL.md`, `README.md`) explicitly states it "only reads local files the user has permission to access" and "does not bypass any legitimate file access control," the capability to read such a broad range of sensitive files, if invoked by an AI agent with elevated privileges or tricked by prompt injection, presents a significant data exfiltration risk. There is no explicit malicious intent (e.g., exfiltration to a remote server) within the code itself, but the broad file access capability makes it a high-risk vulnerability. ```json { "classification": "suspicious", "summary": "The `read_file.py` script is designed to read local file content and includes an extremely broad list of `TEXT_EXTENSIONS` that covers many sensitive file types (e.g., .env, .pem, .key, .ssh, .aws, .history). While the code itself does not contain explicit data exfiltration logic or malicious commands,