Orchestrates multi-agent AI systems with task delegation, agent communication, shared memory, and workflow coordination. Use when users request "multi-agent system", "agent orchestration", "AI agents", "agent coordination", or "autonomous agents".
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Orchestrates multi-agent AI systems with task delegation, agent communication, shared memory, and workflow coordination. Use when users request "multi-agent system", "agent orchestration", "AI agents", "agent coordination", or "autonomous agents".
AI Agent Orchestrator
Build coordinated multi-agent systems for complex task automation.
// agents/specialists.tsimport { Agent, AgentConfig } from'./base';
exportconstResearchAgent = newAgent({
name: 'researcher',
role: 'Research Specialist',
systemPrompt: `You are a research specialist. Your job is to:
- Search for and gather relevant information
- Analyze sources and extract key insights
- Summarize findings clearly
- Cite sources when possible
When you have gathered sufficient information, include [TASK_COMPLETE] in your response.
If you need help from another agent, specify: [HANDOFF:agent_name]`,
tools: [searchTool, webScrapeTool],
});
exportconstWriterAgent = newAgent({
name: 'writer',
role: 'Content Writer',
systemPrompt: `You are a professional content writer. Your job is to:
- Create engaging, well-structured content
- Adapt tone and style to the target audience
- Incorporate research and data effectively
- Edit and refine for clarity
Use the research provided to create compelling content.
When complete, include [TASK_COMPLETE].`,
});
exportconstReviewerAgent = newAgent({
name: 'reviewer',
role: 'Quality Reviewer',
systemPrompt: `You are a quality reviewer. Your job is to:
- Review content for accuracy and clarity
- Check for errors and inconsistencies
- Suggest improvements
- Approve or request revisions
Provide specific feedback. If approved, include [APPROVED].
If revisions needed, include [REVISIONS_NEEDED] with specific changes.`,
});
exportconstPlannerAgent = newAgent({
name: 'planner',
role: 'Task Planner',
systemPrompt: `You are a task planner. Your job is to:
- Break down complex tasks into subtasks
- Identify which specialist agent should handle each subtask
- Create an execution order
- Track progress
Output a structured plan in JSON format:
{
"goal": "...",
"steps": [
{ "step": 1, "agent": "researcher", "task": "..." },
{ "step": 2, "agent": "writer", "task": "..." }
]
}`,
});
Orchestrator
Simple Sequential Orchestrator
// orchestrator/sequential.tsimport { Agent } from'../agents/base';
interfaceWorkflowStep {
agent: Agent;
task: string;
inputFrom?: string;
}
exportclassSequentialOrchestrator {
privateagents: Map<string, Agent> = newMap();
privateresults: Map<string, string> = newMap();
registerAgent(name: string, agent: Agent) {
this.agents.set(name, agent);
}
asyncexecute(workflow: WorkflowStep[]): Promise<Record<string, string>> {
for (const step of workflow) {
const agent = step.agent;
// Get input from previous step if specifiedlet input = step.task;
if (step.inputFrom && this.results.has(step.inputFrom)) {
input = `${step.task}\n\nPrevious output:\n${this.results.get(step.inputFrom)}`;
}
console.log(`Executing: ${agent.name} - ${step.task}`);
const result = await agent.execute(input);
this.results.set(agent.name, result.content);
console.log(`Completed: ${agent.name}`);
}
returnObject.fromEntries(this.results);
}
}
// Usageconst orchestrator = newSequentialOrchestrator();
orchestrator.registerAgent('researcher', ResearchAgent);
orchestrator.registerAgent('writer', WriterAgent);
orchestrator.registerAgent('reviewer', ReviewerAgent);
const results = await orchestrator.execute([
{ agent: ResearchAgent, task: 'Research the latest AI trends in 2024' },
{ agent: WriterAgent, task: 'Write a blog post about AI trends', inputFrom: 'researcher' },
{ agent: ReviewerAgent, task: 'Review the blog post', inputFrom: 'writer' },
]);