notion-cli
Fail
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill's installation process uses a piped bash command (curl -fsSL https://ntn.dev | bash) to execute a script from a remote server. This method lacks verification and execution of unverified scripts from non-trusted domains poses a high security risk.
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with Notion content using ntn pages get, which retrieves data from an external source. This data is untrusted and could be used to inject malicious instructions intended to manipulate agent behavior.
- Ingestion points: Notion page data retrieved via ntn pages get in SKILL.md.
- Boundary markers: No delimiters or safety instructions are present to prevent the agent from following instructions within the data.
- Capability inventory: The skill can manage workers, upload files, and perform direct API operations via the ntn toolset.
- Sanitization: No content filtering or validation is performed on the data retrieved from Notion.
- [COMMAND_EXECUTION]: The skill facilitates the execution of numerous CLI commands that interact with the Notion API and manage workers, granting significant control over external infrastructure that could be abused if the agent is compromised.
Recommendations
- HIGH: Downloads and executes remote code from: https://ntn.dev - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata