forward-message

Warn

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill requires access to sensitive Telegram session files located at ~/.local/share/gramjs/. These files contain credentials for the user's Telegram account. Accessing these paths exposes highly sensitive data to the agent context.
  • [INDIRECT_PROMPT_INJECTION]: The skill implements a workflow that reads external content from Telegram chats using the read command. This represents an ingestion point for untrusted data that could contain malicious instructions.
  • Ingestion points: Telegram message content retrieved via bun "$SCRIPT" read.
  • Boundary markers: The instructions lack delimiters or safety warnings for the agent when processing the content of these messages.
  • Capability inventory: The skill has the ability to execute shell commands via the Bash tool.
  • Sanitization: There is no evidence of sanitization or filtering of message content before it is processed by the agent.
  • [COMMAND_EXECUTION]: The skill executes shell commands by piping a heredoc into the Bash tool to run a TypeScript CLI tool with bun. This execution pattern depends on the integrity of the local tg-cli.ts script.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 7, 2026, 02:12 AM