telegram-inbound-run

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Detected an indirect prompt injection attack surface where untrusted data from external sources is ingested into the agent context.
  • Ingestion points: The skill extracts inbound Telegram messages from the last 24 hours for processing (SKILL.md).
  • Boundary markers: No delimiters or instructions to ignore embedded commands are specified in the documentation for handling the message content.
  • Capability inventory: The agent has the capability to write to local CRM CSV files and execute git commits based on the processed data.
  • Sanitization: No sanitization or filtering logic is documented for the incoming Telegram message content before it is passed to the classification model.
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute local scripts and standard shell utilities.
  • Usage of python3 telegram_inbound_agent.py and tg_auth.py for core message processing and authentication.
  • Usage of diagnostic commands such as grep and cat to inspect local CRM records and log files.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 02:23 PM
Security Audit — agent-trust-hub — telegram-inbound-run