nextflow

Pass

Audited by Gen Agent Trust Hub on Jun 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the Nextflow installation script from the official domain https://get.nextflow.io and provides instructions for installing nf-core from public package registries.
  • [REMOTE_CODE_EXECUTION]: Provides the standard installation command for Nextflow which involves piping a script from https://get.nextflow.io to bash for execution.
  • [COMMAND_EXECUTION]: Includes instructions to use sudo for moving the Nextflow binary to /usr/local/bin/ to facilitate system-wide access.
  • [PROMPT_INJECTION]: The skill manages external data processing which presents a surface for indirect prompt injection:
  • Ingestion points: Processes external CSV samplesheets and configuration files as described in references/running-pipelines.md and references/language.md.
  • Boundary markers: Relies on Nextflow's internal task isolation; no explicit prompt-level delimiters are specified for the agent instructions.
  • Capability inventory: Enables execution of shell and Python scripts within process blocks, local file system access for task execution, and network access for downloading pipelines via nextflow run.
  • Sanitization: Recommends the use of the nf-schema plugin for validating parameters and input data schemas.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 15, 2026, 12:02 AM
Security Audit — agent-trust-hub — nextflow