parallel-processing-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPRIVILEGE_ESCALATIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill offers numerous templates for parallelizing tasks using GNU Parallel,
xargs -P, and background shell jobs. - [EXTERNAL_DOWNLOADS]: Fetches the GNU Parallel source distribution from the official GNU Project repository at
ftp.gnu.orgfor installation. - [PRIVILEGE_ESCALATION]: Provides instructions for installing software using
sudo apt-getandsudo make install, which is standard procedure for system-level tool installation. - [DYNAMIC_EXECUTION]: The "Promise-Like Pattern" reference script uses
evalto dynamically assign process IDs and file paths to variables based on user-provided names. - [INDIRECT_PROMPT_INJECTION]: The skill describes methods for processing untrusted data sources which could theoretically contain malicious content.
- Ingestion points: Data is ingested from external files (
*.txt,urls.txt), TSV/CSV files, and standard input streams. - Boundary markers: Command templates do not include specific delimiters or instructions to ignore potential commands embedded in the processed data.
- Capability inventory: The skill utilizes shell execution and network tools (
wget,curl) across multiple parallel workers. - Sanitization: No explicit sanitization or input validation is demonstrated in the provided code patterns.
Audit Metadata