tooluniverse-antibody-engineering

Pass

Audited by Gen Agent Trust Hub on May 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates data retrieval from established scientific and biological databases, including UniProt, PubMed, IEDB (Immune Epitope Database), and IMGT (International ImMunoGeneTics information system). These interactions are consistent with the skill's stated purpose of antibody engineering and optimization.
  • [COMMAND_EXECUTION]: The documentation provides illustrative Python code snippets for processing biological sequences and interacting with the tooluniverse library. These snippets utilize standard data processing libraries like numpy and do not include any calls to execute arbitrary system commands, perform unauthorized file operations, or spawn dangerous subprocesses.
  • [SAFE]: The skill adopts secure configuration practices by utilizing a .env.template file for managing API keys, ensuring that sensitive credentials are not hardcoded. The provided antibody sequences in the examples are valid biological data with no evidence of embedded malicious instructions, acrostics, or obfuscation. While the skill processes user-supplied sequences which could serve as an indirect prompt injection vector, its capabilities are restricted to scientific analysis and tool calls with no high-risk persistence or privilege escalation mechanisms.
Audit Metadata
Risk Level
SAFE
Analyzed
May 29, 2026, 05:25 AM
Security Audit — agent-trust-hub — tooluniverse-antibody-engineering