Parallelize a single instruction into isolated threads with dedicated workspaces.
npx skills add https://github.com/huangwb8/skills --skill parallel-vibe
Please help me install this Agent Skill. Skill: parallel-vibe Source: https://github.com/huangwb8/skills/tree/main/parallel-vibe Command: npx skills add https://github.com/huangwb8/skills --skill parallel-vibe
Parallelize a single instruction into multiple isolated threads with dedicated workspaces to prevent cross-contamination of results.
Run the parallel-vibe script with a prompt to generate multi-threaded workspaces and review the summary report.
High-intent search queries and answers about installing and using this skill.
Workspace isolation prevents cross-contamination by assigning each parallel thread a dedicated directory, ensuring that planning, implementation, testing, and review outputs remain clean and reproducible without interfering with each other.
Workspace isolation is a mechanism that parallelizes a single instruction into multiple isolated threads with dedicated directories, preventing cross-contamination of results across planning, implementation, testing, and reviewing perspectives.
Yes, you can customize multi-threaded task execution using plan.json to drive continuation and optional synthesis, enforcing per-thread prompts and worker isolation while generating a final summary report across multiple perspectives or models.
To synthesize results from multiple parallel threads, the script enforces per-thread prompts and worker isolation, then aggregates the outputs into a final summary report that integrates planning, implementation, testing, and review perspectives.
Yes, parallel execution supports planning, implementing, testing, and reviewing a single task across multiple perspectives or models, applying optional parallel execution with per-thread prompts and a final summary report.
You should use parallel isolated threads when you need to process a single instruction across multiple perspectives or models simultaneously, requiring clean reproducible outputs and a synthesized final summary report without cross-contamination.
surus-latGenerate synthetic benchmark datasets as JSONL input-output pairs from task specifications.
lossyrobReview PRs across multiple AI models and synthesize independent assessments.
garden-coRun structured cronometro benchmarks for Jazz packages with worker-thread isolation.
synqingDispatch fresh subagents per task for isolated implementation and code review.
