vibe-notionbot

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructions suggest downloading and executing the vibe-notion package from the npm registry at runtime using package runners like npx, bunx, or pnpm dlx if the tool is not locally available.
  • [DATA_EXFILTRATION]: The skill provides commands that read content from the local file system and transmit it to Notion's servers, such as --markdown-file for page creation and the block upload command for image and document uploads.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from Notion workspaces which may contain malicious instructions targeting the AI agent.
  • Ingestion points: Data is ingested via page get, block children, search, and database query commands which retrieve content from external Notion environments (SKILL.md).
  • Boundary markers: There are no defined boundary markers or specific instructions to the agent to disregard instructions embedded within the retrieved Notion content.
  • Capability inventory: The skill possesses the capability to execute a wide range of Notion operations via the vibe-notionbot CLI using Bash, including creating, updating, and deleting pages and databases.
  • Sanitization: No evidence of sanitization, validation, or escaping of retrieved Notion content is present in the instructions before the data is processed by the agent.
  • [COMMAND_EXECUTION]: All functionality in the skill is implemented via shell command execution of the vibe-notionbot CLI tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 02:06 AM
Security Audit — agent-trust-hub — vibe-notionbot