The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill's installation instructions suggest using curl -fsSL https://ntn.dev | bash. This pattern involves piping a remote script directly into a shell interpreter, allowing for arbitrary code execution without integrity verification or user review of the script content.\n- [EXTERNAL_DOWNLOADS]: The skill depends on the external ntn CLI tool, which it suggests downloading from https://ntn.dev or installing via the npm registry. These sources are external to the controlled environment and represent a dependency on third-party infrastructure.\n- [COMMAND_EXECUTION]: The skill operates by executing shell commands using the ntn CLI tool. This provides a broad capability for the agent to interact with the host system and the Notion API, which could be abused if the agent is misled.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests and processes untrusted data from external Notion workspaces.\n
Ingestion points: Data enters the context via ntn pages get, ntn api v1/data_sources/<ds-id>/query, and ntn api v1/search which retrieve user-controlled text from Notion.\n
Boundary markers: No delimiters or instructions are provided to the agent to treat the retrieved content as data rather than instructions.\n
Capability inventory: The skill possesses significant capabilities including reading/writing pages, managing comments, and uploading files.\n
Sanitization: There are no mechanisms described to sanitize or validate the content retrieved from Notion before the agent processes it.

notion