alterlab-datacommons
Warn
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation and reference files (SKILL.md, references/getting_started.md) instruct users to install the
datacommons-clientpackage. This package is not the official Data Commons Python library (which is nameddatacommons), yet the skill provides links to the official Data Commons GitHub repository (github.com/datacommonsorg/api-python). This discrepancy represents a significant supply chain risk, as it encourages the installation of a similarly-named but unverified third-party package. - [PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data from the Data Commons knowledge graph (resolving names, fetching observations, and exploring nodes). This data is processed without adequate isolation or validation, posing a risk of indirect prompt injection.
- Ingestion points: Data is retrieved from the Data Commons API via the Resolve, Node, and Observation endpoints as documented in the
references/directory. - Boundary markers: No delimiters or instructions are used to distinguish untrusted external data from the agent's core instructions.
- Capability inventory: The skill enables the agent to execute Python code, use the Pandas library for data manipulation, and perform network requests for data fetching.
- Sanitization: No evidence of data validation, filtering, or escaping is present in the skill's logic before the retrieved data is used by the agent.
Audit Metadata