notion-master

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly instructs the agent to ask the user to paste their Notion API key and then write that key verbatim into a .env file (and shows example lines containing the secret), which requires the LLM to handle and output secret values directly.

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

• Third-party content exposure detected (high risk: 0.90). The skill suite explicitly calls Notion APIs (e.g., discover_databases.py, search_skill_database.py, manage_blocks.py, manage_comments.py, download_skill.py and create_page.py) to retrieve pages, blocks, comments and attached/external file URLs — i.e., user-generated Notion content and external links that the agent reads and processes at runtime, exposing it to untrusted third-party input that could carry indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skills download skill bundles at runtime by calling the Notion API (e.g. GET https://api.notion.com/v1/pages/{page_id} and extracting the page["properties"]["Skill"]["files"][0]["file"]["url"]) which may point to external file URLs (e.g. https://github.com/...) and those downloaded .skill/.zip/.json/.txt bundles can contain SKILL.md or code that the agent imports and thus directly control prompts or execute code.