bio-assembly-qc

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [SAFE] (SAFE): No malicious instructions, prompt injections, or obfuscated content were detected in the instructions or bibliographic summaries.
  • [NO_CODE] (SAFE): The skill consists entirely of documentation and references; it contains no automated scripts or binaries.
  • [PROMPT_INJECTION] (LOW): The skill possesses an indirect prompt injection surface (Category 8). 1. Ingestion points: The agent processes external sequencing reads and the resulting QUAST reports (HTML/TSV). 2. Boundary markers: No delimiters or warnings are used to separate tool output from instructions. 3. Capability inventory: The agent has terminal access to execute assembly tools. 4. Sanitization: No validation or filtering is applied to the tool output before summarization. This creates a surface where malicious input data could influence agent behavior via tool reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 05:19 PM