dailybot

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). High indirect prompt-injection risk: the dailybot-health workflow fetches pending_messages from Dailybot (third-party human/agent free text) via CLI/HTTP and then incorporates/summarizes that content into the agent’s LLM context for the developer (“You have X messages… ”).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs fetching and executing a remote installer at runtime (e.g., curl -sSL https://cli.dailybot.com/install.sh | bash and irm https://cli.dailybot.com/install.ps1 | iex), which runs remote code and is required to install the dailybot-cli dependency the skill relies on.