bitrix24-workspace

Pass

Audited by Gen Agent Trust Hub on Nov 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill operates by executing a local Node.js CLI script (scripts/b24.cjs) to fetch data from the Bitrix24 API. This is the primary intended function and is restricted to read-only methods.
  • [INDIRECT_PROMPT_INJECTION]: The skill explicitly addresses the risk of malicious instructions embedded in retrieved workspace data (chats, emails, files). It provides mandatory safety rules instructing the agent to treat all retrieved content strictly as data and never as instructions to be executed.
  • [CREDENTIALS_SAFE]: Authentication is managed securely via an inbound webhook URL. The skill provides clear instructions for the user to store this secret in a local configuration file with restricted permissions (chmod 600) or an environment variable, explicitly warning against sharing the credential within the AI chat interface.
  • [SAFE]: Static analysis flagged a destructive system command (rm -rf) in evals/evals.json. Investigation confirms this string is part of a security evaluation case designed to test the agent's resistance to prompt injection. It is not an instruction the skill itself attempts to execute.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 13, 2026, 10:01 PM