dispatching-parallel-agents

por openai

Úsalo cuando enfrentes 2 o más tareas independientes que puedan trabajarse sin estado compartido ni dependencias secuenciales.

npx skills add https://github.com/openai/plugins --skill dispatching-parallel-agents

Dispatching Parallel Agents

Overview

You delegate tasks to specialized agents with isolated context. By precisely crafting their instructions and context, you ensure they stay focused and succeed at their task. They should never inherit your session's context or history — you construct exactly what they need. This also preserves your own context for coordination work.

When you have multiple unrelated failures (different test files, different subsystems, different bugs), investigating them sequentially wastes time. Each investigation is independent and can happen in parallel.

Core principle: Dispatch one agent per independent problem domain. Let them work concurrently.

When to Use

digraph when_to_use {
    "Multiple failures?" [shape=diamond];
    "Are they independent?" [shape=diamond];
    "Single agent investigates all" [shape=box];
    "One agent per problem domain" [shape=box];
    "Can they work in parallel?" [shape=diamond];
    "Sequential agents" [shape=box];
    "Parallel dispatch" [shape=box];

    "Multiple failures?" -> "Are they independent?" [label="yes"];
    "Are they independent?" -> "Single agent investigates all" [label="no - related"];
    "Are they independent?" -> "Can they work in parallel?" [label="yes"];
    "Can they work in parallel?" -> "Parallel dispatch" [label="yes"];
    "Can they work in parallel?" -> "Sequential agents" [label="no - shared state"];
}

Use when:

  • 3+ test files failing with different root causes
  • Multiple subsystems broken independently
  • Each problem can be understood without context from others
  • No shared state between investigations

Don't use when:

  • Failures are related (fix one might fix others)
  • Need to understand full system state
  • Agents would interfere with each other

The Pattern

1. Identify Independent Domains

Group failures by what's broken:

  • File A tests: Tool approval flow
  • File B tests: Batch completion behavior
  • File C tests: Abort functionality

Each domain is independent - fixing tool approval doesn't affect abort tests.

2. Create Focused Agent Tasks

Each agent gets:

  • Specific scope: One test file or subsystem
  • Clear goal: Make these tests pass
  • Constraints: Don't change other code
  • Expected output: Summary of what you found and fixed

3. Dispatch in Parallel

// In Claude Code / AI environment
Task("Fix agent-tool-abort.test.ts failures")
Task("Fix batch-completion-behavior.test.ts failures")
Task("Fix tool-approval-race-conditions.test.ts failures")
// All three run concurrently

4. Review and Integrate

When agents return:

  • Read each summary
  • Verify fixes don't conflict
  • Run full test suite
  • Integrate all changes

Agent Prompt Structure

Good agent prompts are:

  1. Focused - One clear problem domain
  2. Self-contained - All context needed to understand the problem
  3. Specific about output - What should the agent return?
Fix the 3 failing tests in src/agents/agent-tool-abort.test.ts:

1. "should abort tool with partial output capture" - expects 'interrupted at' in message
2. "should handle mixed completed and aborted tools" - fast tool aborted instead of completed
3. "should properly track pendingToolCount" - expects 3 results but gets 0

These are timing/race condition issues. Your task:

1. Read the test file and understand what each test verifies
2. Identify root cause - timing issues or actual bugs?
3. Fix by:
   - Replacing arbitrary timeouts with event-based waiting
   - Fixing bugs in abort implementation if found
   - Adjusting test expectations if testing changed behavior

Do NOT just increase timeouts - find the real issue.

Return: Summary of what you found and what you fixed.

Common Mistakes

❌ Too broad: "Fix all the tests" - agent gets lost ✅ Specific: "Fix agent-tool-abort.test.ts" - focused scope

❌ No context: "Fix the race condition" - agent doesn't know where ✅ Context: Paste the error messages and test names

❌ No constraints: Agent might refactor everything ✅ Constraints: "Do NOT change production code" or "Fix tests only"

❌ Vague output: "Fix it" - you don't know what changed ✅ Specific: "Return summary of root cause and changes"

When NOT to Use

Related failures: Fixing one might fix others - investigate together first Need full context: Understanding requires seeing entire system Exploratory debugging: You don't know what's broken yet Shared state: Agents would interfere (editing same files, using same resources)

Real Example from Session

Scenario: 6 test failures across 3 files after major refactoring

Failures:

  • agent-tool-abort.test.ts: 3 failures (timing issues)
  • batch-completion-behavior.test.ts: 2 failures (tools not executing)
  • tool-approval-race-conditions.test.ts: 1 failure (execution count = 0)

Decision: Independent domains - abort logic separate from batch completion separate from race conditions

Dispatch:

Agent 1 → Fix agent-tool-abort.test.ts
Agent 2 → Fix batch-completion-behavior.test.ts
Agent 3 → Fix tool-approval-race-conditions.test.ts

Results:

  • Agent 1: Replaced timeouts with event-based waiting
  • Agent 2: Fixed event structure bug (threadId in wrong place)
  • Agent 3: Added wait for async tool execution to complete

Integration: All fixes independent, no conflicts, full suite green

Time saved: 3 problems solved in parallel vs sequentially

Key Benefits

  1. Parallelization - Multiple investigations happen simultaneously
  2. Focus - Each agent has narrow scope, less context to track
  3. Independence - Agents don't interfere with each other
  4. Speed - 3 problems solved in time of 1

Verification

After agents return:

  1. Review each summary - Understand what changed
  2. Check for conflicts - Did agents edit same code?
  3. Run full suite - Verify all fixes work together
  4. Spot check - Agents can make systematic errors

Real-World Impact

From debugging session (2025-10-03):

  • 6 failures across 3 files
  • 3 agents dispatched in parallel
  • All investigations completed concurrently
  • All fixes integrated successfully
  • Zero conflicts between agent changes

Más skills de openai

user-context
openai
Cargar o gestionar las preferencias de enrutamiento de origen duraderas del complemento de Análisis de Datos, la lógica de incorporación, el progreso de configuración y el registro de la capa semántica.
official
notion-research-documentation
openai
Investiga contenido en Notion y sintetízalo en informes estructurados, resúmenes o comparaciones con citas. Busca y obtén páginas de Notion mediante consultas específicas, luego organiza los hallazgos por tema con citas en línea de las fuentes y una sección de referencias. Elige entre cuatro formatos de salida (resumen rápido, síntesis de investigación, comparación, informe completo) según el alcance y el objetivo del usuario. Crea y actualiza páginas de Notion usando plantillas integradas; enlaza las fuentes directamente y registra los cambios a medida que llega nueva información...
official
rcsb-pdb-skill
openai
Envía solicitudes compactas de RCSB PDB para metadatos básicos, consultas de la API de búsqueda y descargas FASTA. Úsalo cuando un usuario desee resúmenes concisos de RCSB; guarda el JSON sin procesar o…
official
pdf
openai
Lectura, creación y validación de PDF con renderizado visual y generación programática. Renderiza páginas PDF a PNG para inspección visual del diseño, espaciado y tipografía antes de la entrega usando Poppler (pdftoppm). Genera PDFs programáticamente con reportlab para un formateo confiable; extrae texto y metadatos con pdfplumber o pypdf. Aplica estándares de calidad: sin texto recortado, elementos superpuestos, tablas rotas ni artefactos de renderizado; solo guiones ASCII, citas legibles para humanos. Usa...
official
test-coverage-improver
openai
Improve test coverage in the OpenAI Agents JS monorepo: run `pnpm test:coverage`, inspect coverage artifacts, identify low-coverage files and branches, propose…
official
playwright
openai
Automatización de navegador desde terminal con instantáneas de elementos y flujos de trabajo interactivos de interfaz de usuario. Opera mediante un script envolvente de playwright-cli (requiere npx); admite modos headless y con interfaz gráfica para depuración visual. Flujo de trabajo principal: abrir página, tomar instantánea para referencias estables de elementos, interactuar usando referencias, volver a tomar instantánea tras navegación o cambios en el DOM. Incluye llenado de formularios, clics, escritura, gestión de múltiples pestañas, captura de pantalla/PDF y grabación de trazas para depuración de flujos. Las referencias de elementos (por ejemplo, e3, e15)...
official
ukb-topmed-phewas-skill
openai
Obtén resúmenes PheWAS compactos de UKB-TOPMed para variantes individuales aceptando entrada de rsID, GRCh37 o GRCh38 y resolviendo a la consulta GRCh38 requerida. Úsalo cuando un…
official
code-review-context
openai
Contexto visible del modelo
official