**Instructions:**

**Code Selection**

1. **Identify Relevant Codes**: For each excerpt, determine all applicable codes from the provided list that accurately describe the content of the excerpt.

2. **Utilize Metadata for Context**: Leverage the provided metadata to gain a comprehensive understanding of each excerpt, ensuring accurate code assignments.

3. **Handle Non-Applicable Excerpts**: If none of the provided codes are relevant to a current excerpt, assign the code "VOID" in the `"code_name"` field and supply an appropriate justification.

**Provide Justifications**

4. **Explain Code Assignments**: For every assigned code, include a 2-3 sentence justification detailing why the code is pertinent to the **current excerpt** (not the context).

5. **Justify "VOID" Assignments**: If "VOID" is assigned, provide a 2-3 sentence explanation clarifying why none of the provided codes are suitable for the **current excerpt**.

**Formatting Requirements**

6. **JSON Structure**: Format the output as a JSON object containing an `"assignments"` array. Each object within this array corresponds to a single excerpt and should include the following keys:

   - `"unique_id"`: The unique identifier of the excerpt.
   
   - `"codeList"`: An array of code assignments, each comprising:
     - `"code_name"`: The assigned code.
     - `"code_justification"`: The rationale for the code assignment.

7. **Maintain Order**: Ensure that each JSON object aligns with the order of the provided excerpts.

**Assigning Multiple Codes**

8. **Comprehensive Coding**: Assign all applicable codes to each excerpt; some excerpts may warrant multiple codes.

9. **Use Defined Codes Only**: Restrict assignments to codes explicitly defined in the provided list.

**Ensure JSON Validity**

10. **Accurate Formatting**: Construct the JSON meticulously to avoid syntax errors such as missing commas, brackets, or braces.

11. **No Extraneous Text**: Do not include any additional text, explanations, or comments outside the JSON code block.

**Example Output Format:**

```json
{
  "assignments": [
    {
      "unique_id": 1,
      "codeList": [
        {
          "code_name": "Sample_Code1",
          "code_justification": "This code was applied because the excerpt discusses..."
        },
        {
          "code_name": "Sample_Code2",
          "code_justification": "This code was applied because the excerpt also relates to..."
        }
      ]
    },
    {
      "unique_id": 2,
      "codeList": [
        {
          "code_name": "VOID",
          "code_justification": "No code was applicable as the excerpt does not align with any provided categories."
        }
      ]
    }
  ]
}

Contextual Metadata:

You will receive a JSON object containing all relevant metadata for each current excerpt. Use this metadata exclusively as context to inform your code assignments. Do not include any metadata in your output.

You will also receive context from the conversation in which the current excerpt took place. Do NOT code for this surrounding context; ONLY code for the current excerpt.

Output:

Submit your response as a JSON object containing only the assignments with the assigned codes for each excerpt – do not add commentary or any non-JSON text.

