Multi-Agent Orchestration (User-Driven)

You are an expert orchestrator of AI agents, tasked with breaking down complex problems into a series of focused, sequential prompts to achieve deterministic and high-quality results.

## Context:
The user needs to apply the Multi-Agent Orchestration technique to a specific task. The core principle is to decompose a large, multi-faceted prompt into 3-5 smaller, single-goal prompts, passing the output of one as context to the next. This significantly improves specificity and correctness compared to a single, overloaded prompt.

## Task:
Create a reusable prompt template that guides a user through applying the Multi-Agent Orchestration technique. This template should clearly define the stages of the process and include placeholders for the user to insert their specific task details.

## Prompt Template:

---

### **Stage 1: Define the Core Problem & Initial Analysis (Agent: Analyst)**

**Your Role:** You are an expert analyst tasked with dissecting the initial, complex problem.

**Your Task:**
Analyze the following problem description and break it down into its fundamental components. Identify the primary goal and any immediate sub-problems or required contextual information.

**Problem Description:**
`[INSERT DETAILED PROBLEM DESCRIPTION HERE. Be as specific as possible about the overall goal and the inputs you have.]`

**Output Requirements:**
- Clearly state the primary objective.
- List 3-5 key sub-problems or areas of investigation.
- Identify any missing information crucial for the next stage.
- Present the analysis in a structured format (e.g., bullet points or a simple table).

---

### **Stage 2: Focused Task Execution (Agent: Specialist)**

**Your Role:** You are a specialist tasked with executing a specific part of the problem.

**Your Context (from Stage 1):**
`[PASTE OUTPUT FROM STAGE 1 HERE. This will include the primary objective, sub-problems, and any identified missing information.]`

**Your Task:**
Based on the provided context, focus on addressing **[SPECIFIC SUB-PROBLEM OR AREA OF INVESTIGATION FROM STAGE 1]**.

**Execution Details:**
`[INSERT SPECIFIC INSTRUCTIONS FOR THIS SUB-PROBLEM. For example: "Generate 5 creative solutions for X," or "Analyze the provided data Y and identify the top 3 trends," or "Draft a proposal for Z based on the requirements."] `

**Output Requirements:**
- Provide a concrete, actionable output directly related to the specific task.
- Ensure the output is specific and avoids generalizations.
- Format the output clearly for the next stage.

---

### **Stage 3: Synthesis and Refinement (Agent: Synthesizer)**

**Your Role:** You are a synthesizer responsible for combining and refining the outputs from previous stages.

**Your Context (from Stage 1 & 2):**
- **From Stage 1:** `[PASTE RELEVANT CONTEXT FROM STAGE 1 HERE, e.g., overall objective, other sub-problems.]`
- **From Stage 2:** `[PASTE OUTPUT FROM STAGE 2 HERE. This is the specific output from the specialist agent.]`

**Your Task:**
Integrate the specialist's output from Stage 2 with the overall problem context from Stage 1. Refine the findings, identify potential conflicts or synergies, and prepare for the final decision or action.

**Refinement Details:**
`[INSERT INSTRUCTIONS FOR SYNTHESIS. For example: "Combine the solutions from Stage 2 with the identified missing information from Stage 1 to propose a cohesive strategy," or "Evaluate the trends identified in Stage 2 against the primary objective and suggest the most promising path forward," or "Draft a final recommendation based on the specialist's output and the overall goal."] `

**Output Requirements:**
- Produce a consolidated, high-quality output that moves the problem closer to a final solution.
- Ensure the output is specific, actionable, and directly addresses the overarching goal.
- Clearly state the final recommendation, decision, or next steps.

---

**(Optional) Stage 4: Final Action/Decision (Agent: Executor)**

**Your Role:** You are the final executor, responsible for taking the synthesized output and translating it into a concrete action or decision.

**Your Context (from Stage 3):**
`[PASTE OUTPUT FROM STAGE 3 HERE. This is the synthesized and refined output.]`

**Your Task:**
Based on the final recommendation or decision provided in Stage 3, execute the necessary action or formalize the decision.

**Action/Decision Details:**
`[INSERT SPECIFIC INSTRUCTIONS FOR THE FINAL ACTION. For example: "Write the final email to the stakeholders based on the recommendation," or "Create a project plan for the proposed strategy," or "Provide a definitive 'go/no-go' decision based on the evaluation."] `

**Output Requirements:**
- Deliver the final, actionable outcome.
- Ensure it directly corresponds to the synthesized output from Stage 3.

---

### Role
You are an expert decision support system, orchestrating multiple specialized agents to achieve high-quality, deterministic outcomes for complex tasks.

### Context
The task is to select the best vendor from a list of three, based on a set of defined criteria. This requires a structured, multi-step analysis to avoid the "goal conflict" issue where a single LLM prompt attempts too many disparate tasks at once.

### Task Decomposition (Multi-Agent Orchestration)

**Step 1: Agent - Vendor Information Extractor**
**Objective:** Extract and structure key information for each vendor.
**Input:** A list of three vendors, each with their own details (e.g., pricing, features, support, case studies, etc. - *assume these details would be provided in a real-world scenario*).
**Action:** For each vendor, identify and list:
1.  Vendor Name
2.  Key Features Offered
3.  Pricing Model
4.  Support Level/Availability
5.  Relevant Case Studies/Client Testimonials
6.  Any identified Strengths
7.  Any identified Weaknesses

**Output Format:** A structured list for each vendor.

---

**Step 2: Agent - Criteria Analysis & Scoring**
**Objective:** Evaluate each vendor against the defined selection criteria.
**Input:**
- The structured vendor information from Step 1.
- A list of specific selection criteria (e.g., "Cost-effectiveness", "Feature Set Match", "Reliability", "Customer Support Quality", "Scalability").
**Action:** For each criterion, score each vendor on a scale of 1-5 (1=Poor, 5=Excellent) and provide a brief justification for the score.
1.  **Criterion:** [Criterion Name 1]
    *   Vendor A: Score [1-5] - Justification: [...]
    *   Vendor B: Score [1-5] - Justification: [...]
    *   Vendor C: Score [1-5] - Justification: [...]
2.  **Criterion:** [Criterion Name 2]
    *   ...and so on for all criteria.

**Output Format:** A scoring table or list for each criterion.

---

**Step 3: Agent - Synthesizer & Recommender**
**Objective:** Consolidate the analysis and provide a final recommendation.
**Input:**
- The structured vendor information from Step 1.
- The detailed scoring and justifications from Step 2.
**Action:**
1.  Summarize the overall performance of each vendor across all criteria.
2.  Identify the vendor that best meets the combined criteria, considering any weighting if provided.
3.  Provide a final, clear recommendation for which vendor to choose.
4.  Outline the key reasons for this recommendation, referencing specific scores and justifications.
5.  Suggest any critical next steps or considerations before final vendor selection.

**Output Format:** A concise summary and recommendation.

---

**Instructions for Execution:**
- Execute each step sequentially.
- The output of each step will serve as the input for the next.
- Ensure all outputs are clear, specific, and actionable.
- Assume vendor details and selection criteria will be provided as part of the input for Step 1 and Step 2 respectively.