brenda-database
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes potentially untrusted data from an external API.
- Ingestion points: Data is retrieved from the BRENDA database SOAP API (brenda-enzymes.org) via the brenda_client module.
- Boundary markers: There are no explicit delimiters or instructions provided to separate the external database results from the agent's core instructions.
- Capability inventory: The skill has the capability to write various file formats (CSV, JSON, Excel, PNG, and TXT) to the local filesystem using libraries such as pandas and matplotlib.
- Sanitization: While the skill uses regular expressions to extract specific numeric fields, it does not fully sanitize or validate all free-text fields (like commentary) returned by the API.
- [EXTERNAL_DOWNLOADS]: The skill connects to and retrieves biochemical research data from the official BRENDA database, which is a well-known and established scientific information system.
Audit Metadata