alterlab-open-notebook
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The test utility "scripts/test_open_notebook_skill.py" employs the "compile()" function to programmatically verify that the provided example scripts are syntactically correct. This execution is limited to local files within the skill package.
- [EXTERNAL_DOWNLOADS]: Instructions in "SKILL.md" direct the user to download a configuration file from "https://raw.githubusercontent.com/lfnovo/open-notebook/main/docker-compose.yml". This download targets the project's public repository on a well-known service and is a standard procedure for the application's deployment.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it is designed to ingest untrusted data from external sources.
- Ingestion points: External data enters the system through the "/api/sources" endpoint, which supports web URLs and file uploads (found in "SKILL.md" and "scripts/source_ingestion.py").
- Boundary markers: None identified in the provided documentation or scripts to delimit untrusted content.
- Capability inventory: The skill facilitates interaction with multiple AI providers (OpenAI, Anthropic, Google, etc.) and performs search/transformation operations on ingested content.
- Sanitization: No explicit sanitization or filtering of ingested content is documented.
- [SAFE]: The skill identifies and follows data privacy best practices by providing a self-hosted architecture and using local storage for research materials.
Audit Metadata