Episode Orchestrator Pro

Workflow coordination agent that manages the end-to-end podcast episode production pipeline, routing tasks to specialized agents, validating payloads, and consolidating outputs into structured JSON deliverables.

When to Use This Agent

Choose this agent when you need to:

• Orchestrate a multi-agent podcast production pipeline from research through post-production
• Validate episode payloads for completeness before dispatching to downstream specialist agents
• Coordinate sequential and parallel agent invocations while preserving output chain dependencies
• Generate consolidated JSON status reports aggregating results from all pipeline stages

Consider alternatives when:

• You need deep academic research synthesis for a specific episode topic (use the Academic Research Synthesizer Copilot)
• Your task is specifically about guest identification and outreach (use the Guest Outreach Coordinator)

Quick Start

Configuration

Copy
name: episode-orchestrator-pro
type: agent
category: podcast-creator-team

Example Invocation

Copy
claude agent:invoke episode-orchestrator-pro '{"title":"The Future of Edge Computing","duration":"45min","airDate":"2026-04-15","guests":["Dr. Lin Chen"],"topics":["edge-ai","5g-integration","latency-optimization"]}'

Example Output

Copy
{
  "status": "success",
  "episode": "The Future of Edge Computing",
  "pipeline_stages": {
    "research": {"status": "complete", "sources": 28, "briefing_pages": 6},
    "guest_prep": {"status": "complete", "questions": 12, "bio_verified": true},
    "script_draft": {"status": "complete", "segments": 5, "word_count": 4200},
    "audio_notes": {"status": "complete", "music_cues": 3, "ad_slots": 2}
  },
  "next_action": "Host review of script draft before recording session",
  "warnings": ["Guest confirmed for 30min only — reduce segment 4"]
}

Core Concepts

Episode Pipeline Overview

Aspect	Details
Payload Validation	Verify title, duration, airDate, guest list, and topic tags before pipeline dispatch
Agent Sequence	Research -> Guest Prep -> Script Draft -> Audio Planning (configurable order)
Error Strategy	Capture failures per stage; continue pipeline for non-blocking errors, halt on critical
Output Format	Consolidated JSON with per-stage status, warnings, and recommended next actions

Orchestration Flow Architecture

┌─────────────────┐     ┌─────────────────┐
│  Episode Payload │────▶│  Payload        │
│  (JSON input)    │     │  Validator      │
└─────────────────┘     └─────────────────┘
        │                       │
        ▼                       ▼
┌─────────────────┐     ┌─────────────────┐
│  Research Agent  │────▶│  Guest Prep     │
│  (topic deep-   │     │  Agent (bios,   │
│   dive)          │     │  questions)     │
└─────────────────┘     └─────────────────┘
        │                       │
        ▼                       ▼
┌─────────────────┐     ┌─────────────────┐
│  Script Draft    │────▶│  Audio Planning │
│  Agent (segments,│     │  Agent (cues,   │
│  transitions)    │     │  timing)        │
└─────────────────┘     └─────────────────┘

Configuration

Parameter	Type	Default	Description
required_fields	array	["title","duration","airDate"]	Fields that must be present in the episode payload to proceed
agent_sequence	array	["research","guest_prep","script","audio"]	Ordered list of agent stages to invoke
halt_on_failure	boolean	false	When true, stop the entire pipeline if any stage fails; when false, skip and continue
clarification_limit	integer	1	Maximum clarifying questions before attempting to proceed with available information
output_format	string	json	Response format: json for machine consumption or markdown for human-readable reports

Best Practices

1. Validate Payloads Strictly Before Dispatching Catching missing or malformed fields at the orchestration layer prevents cascading failures in downstream agents. Define a clear schema for required fields and validate data types, date formats, and field lengths before invoking the first agent in the sequence. A rejected payload with a clear error message saves far more time than a half-completed pipeline.

2. Preserve Agent Output Chain for Downstream Consumption Each agent's output should be stored and made available to subsequent agents in the sequence. The script draft agent needs research findings; the audio planning agent needs the script segments. Pass the cumulative context forward rather than requiring each agent to re-derive information that a previous agent already produced.

3. Handle Partial Failures Gracefully Not every stage failure should halt the pipeline. If the guest prep agent fails because a guest's biography is unavailable, the script and audio stages can still proceed with placeholder content. Design the orchestrator to classify failures as blocking or non-blocking and continue the pipeline whenever possible.

4. Limit Clarification Loops to Prevent Stalling When a payload is incomplete, ask exactly one targeted clarification question. If the information is still not available after the limit, proceed with reasonable defaults and flag the assumption in the output. Unbounded clarification loops stall the production timeline and frustrate users.

5. Generate Machine-Readable and Human-Readable Outputs The consolidated JSON is essential for downstream automation, but producers also need a human-readable summary they can review quickly. Support both json and markdown output formats so the same orchestration run serves both automated pipelines and editorial review workflows.

Common Issues

1. Downstream agent receives truncated context from previous stage When the research agent produces a lengthy briefing, naive payload forwarding may hit token or message-size limits. Implement a summarization step between stages that condenses the previous agent's output to the essential findings and citations needed by the next agent, preserving critical information while staying within size constraints.

2. Pipeline timing assumptions conflict with guest availability The orchestrator may schedule recording prep before guest confirmation arrives. Build a dependency check that verifies guest confirmation status before invoking the guest prep and script draft stages. When confirmation is pending, queue those stages and proceed with research and audio planning in parallel.

3. JSON output validation fails due to unescaped special characters Agent outputs sometimes contain unescaped quotes, newlines, or Unicode characters that break JSON parsing. Apply a JSON sanitization pass to every agent's raw output before incorporating it into the consolidated response, and include a final JSON.parse() validation step before returning the orchestrator's response.