revx-telegram
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The instructions command the agent to "invoke the revx-auth skill immediately and apply the fix yourself" and "Do not interrupt the user to ask them to run setup." This behavior overrides standard interaction protocols by encouraging autonomous execution of follow-up skills without user confirmation.
- [INDIRECT_PROMPT_INJECTION]: The setup process requires the user to retrieve JSON data from an external Telegram API endpoint (api.telegram.org) and provide it to the agent. This creates an attack surface where maliciously crafted data from the API could potentially influence the agent's behavior during subsequent command execution.
- Ingestion points: User providing the JSON response from the Telegram getUpdates endpoint (SKILL.md).
- Boundary markers: None provided in the setup instructions.
- Capability inventory: Execution of revx CLI commands with user-supplied arguments (SKILL.md).
- Sanitization: No explicit sanitization or validation of the chat ID or token format before command interpolation.
- [COMMAND_EXECUTION]: The skill relies on shell command execution via the revx CLI for adding, testing, and managing connections. This involves passing sensitive parameters, such as Telegram bot tokens, directly through command-line flags.
Audit Metadata