bio-workflow-methods-docwriter
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8).
- Ingestion points: The script
scripts/extract_nextflow_run.pyreads data fromtrace.txtand the contents of shell scripts (e.g.,.command.sh) from the Nextflow work directory. - Boundary markers: Absent. The
templates/methods_report.mdandtemplates/paper_summary.yamlfiles use standard interpolation (e.g.,{{steps_block}}) without delimiters or instructions for the agent to ignore embedded commands within that data. - Capability inventory: The skill uses Python scripts to read/write local files and relies on the AI agent to summarize and format the extracted data into reports.
- Sanitization: The scripts use
yaml.safe_loadandyaml.safe_dumpto prevent YAML-based code execution, but they do not sanitize or filter the natural language content extracted from logs, which could contain malicious instructions meant to subvert the agent's behavior during report generation.
Audit Metadata