skills/reason-machines/data-skills/harvard-artifacts-collection-data-engineering-analytics/Gen Agent Trust Hub
harvard-artifacts-collection-data-engineering-analytics
Warn
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions direct the user to clone a project from a personal GitHub repository (
https://github.com/Manali0711/Harvard-Artifacts-Collection-Data-Engineering-Analytics-App.git). This source is unverified and not associated with a trusted organization or the skill author. - [COMMAND_EXECUTION]: The project requires executing shell commands such as
git clone,pip install, andstreamlit runto download, install dependencies, and run the external application code. - [INDIRECT_PROMPT_INJECTION]: The skill includes an ETL pipeline that ingests artifact metadata from a public API (
https://api.harvardartmuseums.org), which represents an attack surface for indirect instructions hidden in external data. - Ingestion points: External data enters the system through the
fetch_artifacts_from_apifunction inSKILL.md. - Boundary markers: The code lacks boundary markers or explicit instructions to ignore potentially malicious commands embedded in the fetched JSON records.
- Capability inventory: The skill possesses network access (
requests), database interaction capabilities (mysql-connector-python), and data visualization tools (streamlit,plotly). - Sanitization: There is no evidence of sanitization or filtering of the API content beyond basic string truncation during the transformation step.
Audit Metadata