telnetshell

Fail

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEDATA_EXFILTRATIONPERSISTENCEPRIVILEGE_ESCALATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill provides scripts (such as enum_security.txt) and instructions to locate and read highly sensitive credential files, including /etc/shadow, private SSH keys (id_rsa, .pem, .key), and files containing hardcoded passwords.
  • [DATA_EXFILTRATION]: Documentation in SKILL.md and examples.md describes techniques for exfiltrating data from the target device, such as piping firmware dumps to a remote listener using nc or serving files via a temporary busybox httpd instance.
  • [PERSISTENCE]: The skill contains explicit instructions for maintaining access to target devices by adding SSH public keys to /root/.ssh/authorized_keys and injecting backdoor commands into system startup scripts like /etc/rc.local or /etc/init.d/rcS.
  • [PRIVILEGE_ESCALATION]: The skill includes commands to identify potential privilege escalation vectors, such as searching for SUID/SGID binaries and testing for sudo permissions on the remote target.
  • [COMMAND_EXECUTION]: The provided telnet_helper.py script uses the pexpect library to spawn and control the system telnet client, allowing the agent to execute arbitrary shell commands on remote hosts.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted output from remote telnet shells directly into the agent's context, creating a vulnerability where a malicious or compromised device could influence the agent's behavior via crafted terminal output.
  • Ingestion points: Shell output is read by telnet_helper.py via pexpect and returned to the agent.
  • Boundary markers: None identified; the output is processed as plain text.
  • Capability inventory: The agent can execute commands, read/write files, and perform network operations on the host and the target device.
  • Sanitization: The helper script cleans ANSI codes but does not perform safety validation or escaping on the content of the shell responses.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of external dependencies, specifically the pexpect Python library and system-level inetutils for the telnet client.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 16, 2026, 01:29 PM