alterlab-metabolomics-wb
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill performs network requests to the NIH Metabolomics Workbench REST API (
https://www.metabolomicsworkbench.org/rest/). This is a well-known academic repository and the operations are consistent with the skill's stated purpose of retrieving metabolomics study data, standardized nomenclature, and chemical information. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from a REST API, which constitutes a vulnerability surface for indirect prompt injection if the remote source were compromised.
- Ingestion points: The helper script
scripts/query_metabolomics_wb.pyretrieves data from multiple REST endpoints, processing it as either JSON or tab-separated values (TSV). - Boundary markers: There are no explicit delimiters or boundary markers implemented in the script or instructions to isolate the retrieved data from the agent's control flow.
- Capability inventory: The skill is configured with access to
Bash(curl:*),Bash(python:*),WebFetch, andReadtools, providing a pathway for subsequent actions based on the ingested data. - Sanitization: The script performs standard data parsing (JSON decoding and string splitting) but does not include security-specific sanitization or filtering of the content returned by the API.
Audit Metadata