Results for “parallel”

Concurrent investigation of independent failures. Use when multiple unrelated issues need parallel resolution.

The English summary is being prepared. Showing the original description.

Shows you how to use Cowork's parallel workers to run multiple tasks at the same time instead of one after another — dramatically cutting the time on batch work. For processing large batches of files or research tasks, when running multiple independent analyses simultaneously, or during a performance optimization session. Use when the user says "speed up", "parallel workers", "batch task", "process multiple files", "why is cowork so slow", "faster", "do several things at once", or "can cowork multitask".

The English summary is being prepared. Showing the original description.

Use ONLY when the user explicitly asks to orchestrate parallel external agents (Codex or Claude Code) in tmux sessions via tp (tmux-pilot). Or when a user tells you that you're the boss or orchestrator. You can also use this if you are an operclaw agent. NEVER auto-invoke for a generic "do this in parallel" request. For in-process subagents, use superpowers:dispatching-parallel-agents instead.

The English summary is being prepared. Showing the original description.

Coordinate structured thinking and multi-agent parallel execution for complex tasks. Use when tackling multi-step projects, planning parallel work, breaking down complex problems, coordinating specialist tasks, facing architectural decisions, or when user mentions "workflow", "orchestration", "multi-step", "coordinate", "parallel execution", "structured thinking", "break this down", "plan this out", "how should I approach", or needs help planning complex implementations.

The English summary is being prepared. Showing the original description.

Breaks work into ordered, parallel-dispatchable tasks with an execution DAG. Output format is consumable by CI matrices and parallel agent runners — each task has a stable ID, declared file writes, conflict edges, and branch suffix. Use when you have a spec, brief, or shaper output and need to decompose it into implementable units. Use when a task feels too large, when scope spans multi-repo or multi-week work, or when parallel execution across multiple agents is on the table.

The English summary is being prepared. Showing the original description.

The intelligence layer of WinDAGs. Decomposes natural language tasks into Hierarchical Task DAGs (HTDAGs), matches subtasks to skills, executes waves in parallel, and dynamically expands nodes when complexity exceeds executor capability. Use for 'orchestrate', 'execute DAG', 'parallel agents', 'decompose task', 'coordinate skills'. NOT for single-skill tasks or simple linear workflows.

The English summary is being prepared. Showing the original description.

分布式并行策略选型与实测（USP / CP 通信掩盖 / CFG / TP/RSP/PP 概览；含拓扑相关选型 （按实测拓扑分域条件化：域内 bulk vs 跨域 head-parallel 翻转）与 AlltoAllV 缺陷绕过）。在 model-auto-optimization 中承担 S3：优先 USP、结合拓扑带宽差异选 CP，少量 step + 多 rank 验证特性开启与掩盖（产物：并行方案 + 多 rank 证据）。当用户需要多卡并行 策略选择、**序列并行形态抉择**（纯 Ulysses vs 复合 AllGather-KV×Ulysses：按 GQA / 跨域带宽 / 形态 plumbing 条件化定胜负）、**并行 × 稀疏叠加**（seam 契约：先汇聚后稀疏、 窗口偏移、块对齐、per-head 掩码；含「稀疏看似生效实则未生效」判定）、通信掩盖调优 （含**掩盖率上限**：1-1/n 何时成立、c/f 决定的真实上限、没生效的排查）、**并行方案 差异归因**（阶段 Δ 分解 / 集合通信按 communicator 归属 / 4→8 卡线性度），或排查多卡 跑不动 / 通信暴露大 / 换卡组 / 端口 bind / HCCL 带宽验证问题时使用；**并收编原并行作用域诊断**： 改了 SP/CP/Ulysses/AllGather-KV 后**不报错但结果没变/性能没变**、或小规模能跑大规模崩 （如 Ascend EE1003 coreDim 超限）时，用本技能证明"改动到底有没有生效"（判别量逐层收窄 + 两侧对照 + 日志≠生效，见 `references/scope-effectiveness-check.md`）； 即使用户只说 "多卡跑不动""通信暴露大""为什么没达到 6/7 的掩盖率""CP 和 USP 该选哪个""CP 叠稀疏 怎么不生效"而未说并行，也应触发。特性档位/接口事实见 `docs/zh/features/parallelism.md` /`usp.md`（仓内真源），框架侧开启见 framework-integration； 本技能承载选型决策、monkey-patch 掩盖与多卡诊断实测。 由 dev-workflow 多卡场景触发，亦由 model-auto-optimization 的 S3 阶段触发。

The English summary is being prepared. Showing the original description.

Task claiming and ownership tracking for parallel agent execution. Agents claim tasks from a shared queue with file ownership enforcement. Use during parallel execution phases when multiple agents work simultaneously.

The English summary is being prepared. Showing the original description.

Multi-agent collaboration plugin that spawns N parallel subagents competing on the same task via git worktree isolation. Agents work independently, results are evaluated by metric or LLM judge, and the best branch is merged. Use when: user wants multiple approaches tried in parallel — code optimization, content variation, research exploration, or any task that benefits from parallel competition. Requires: a git repo.

The English summary is being prepared. Showing the original description.

Fornece treinamento distribuído de LLM nativo do PyTorch usando torchtitan com paralelismo 4D (FSDP2, TP, PP, CP). Use para fazer pré-treinamento de Llama 3.1, DeepSeek V3 ou modelos customizados em escala de 8 a 512+ GPUs com Float8, torch.compile e checkpoint distribuído.

The English summary is being prepared. Showing the original description.

Treina modelos de linguagem grandes (2B-462B de parâmetros) usando NVIDIA Megatron-Core com estratégias avançadas de paralelismo. Use ao treinar modelos >1B de parâmetros, precisar de máxima eficiência de GPU (47% MFU em H100), ou necessitar de paralelismo tensor/pipeline/sequência/contexto/expert. Framework pronto para produção usado em Nemotron, LLaMA, DeepSeek.

The English summary is being prepared. Showing the original description.

6つの専門レビュアーロールを並列実行し、consensusLevel（複数ロールの合意度）と Tech Lead レポート（top3指摘・blindSpots・consensusSummary）で結果を統合する マルチエージェントレビュー entry skill。 Parallel multi-role review with consensus scoring (consensusLevel) and Tech Lead report. Use when a major release needs exhaustive multi-angle review, or when a single-perspective review is not enough and you want confidence that no reviewer angle was missed（重要リリース前の網羅レビュー・多視点の確証が 欲しいとき）。

The English summary is being prepared. Showing the original description.

Coordinate parallel autonomous operations. Use when running parallel features, managing concurrent work, coordinating multiple agents, or optimizing throughput.

The English summary is being prepared. Showing the original description.

Complete guide for TUnit new-generation testing framework. Use when creating test projects with TUnit or migrating from xUnit to TUnit. Covers Source Generator driven test discovery, AOT compilation support, fluent async assertions. Includes project creation, [Test] attribute, lifecycle management, parallel control, and xUnit syntax comparison. Keywords: TUnit, tunit testing, source generator testing, AOT testing, new generation testing framework, [Test], [Arguments], TUnit.Assertions, Assert.That, Before(Test), After(Test), NotInParallel, TUnit.Templates, Microsoft.Testing.Platform, TUnit vs xUnit, parallel execution

The English summary is being prepared. Showing the original description.

Coordinate multiple AI agents with assigned owners, clear work boundaries, and a Kanban board that tracks progress. Organize separate branches or worktrees, collect test and handoff evidence, and define checks that must pass before combining their work.

• Assign parallel tasks to multiple AI agents
• Track progress and blockers across sessions
• Separate editing work across branches or worktrees

Implement asynchronous work and changing data streams in Android and Kotlin Multiplatform apps using Kotlin Coroutines and Flow. Apply patterns for parallel loading, UI state updates, cancellation, error handling, and tests that check emitted values.

• Loading several data sources in parallel
• Managing screen state with StateFlow
• Handling search input delays and network retries

Add background jobs to Django with Celery, moving work such as email delivery and PDF generation outside web requests. Covers scheduled runs, retries, linked or parallel tasks, monitoring, and tests, with Redis or RabbitMQ carrying task messages.

• Moving email delivery or PDF generation into background jobs
• Scheduling inventory updates and recurring notifications
• Designing retries and safeguards against duplicate processing

Plan autonomous Claude Code workflows, from sequential development steps to parallel agents working from a specification. The material explains how to preserve progress between runs, check quality, coordinate merges, and set limits that stop repeated execution.

• Designing a sequential development pipeline
• Preserving progress across repeated agent runs
• Coordinating parallel work and merges

Coordinate coding projects through Claude DevFleet by splitting work into dependent tasks and dispatching Claude Code agents in separate Git working directories. Monitor progress, manage automatic task sequencing, and review structured reports of changes, tests, and errors.

• Splitting a coding project into dependent tasks
• Running coding agents in separate Git working directories
• Monitoring concurrent development tasks

Turn a large engineering objective into a written plan with steps sized for individual pull requests, or proposed code changes. Each step includes context, dependencies, checks, and completion criteria so another agent can continue the work.

• Splitting a large feature into multiple pull requests
• Planning a migration across several sessions
• Identifying steps agents can work on in parallel

Run machine learning training and predictions on cloud GPUs, then publish models as APIs that other applications can call. The skill guides Python-based setup, automatic scaling, persistent storage, parallel processing, and scheduled jobs.

• Run machine learning jobs without managing GPU servers
• Publish a model as an automatically scaling API
• Process large batches of data in parallel

Delegate feature development, code refactoring, and pull request reviews to OpenAI Codex CLI. The skill also covers monitoring long tasks and fixing multiple issues in parallel using separate Git working directories.

• Adding a feature to an existing project
• Refactoring an authentication module
• Reviewing proposed code changes

Coordinate 15 agents implementing claude-flow v3 across security, core systems, integration, testing, performance, and release work. Provides role assignments, task dependencies, GitHub tracking examples, and a four-phase schedule spanning 14 weeks.

• Assign claude-flow v3 work across 15 agents
• Run tasks in parallel while respecting dependencies
• Track milestones and progress through GitHub issues

Create workflows that pause for approval and resume later, or scripts that distribute work across multiple agents. Choose a saved template for ongoing processes or a JavaScript script to collect structured results from parallel tasks.

• Build a workflow with approval checkpoints
• Split review work across multiple agents
• Collect structured results from research tasks