Show me your Kling prompt templates for technical specification drafting.

Everyone's talking about using Kling for creative briefs and marketing copy, but I've been testing it for drafting technical specs for CRM integrations and custom object builds. The out-of-the-box "technical writer" prompts are too generic. They miss the crucial detail needed to actually build something.

I need prompts that force Kling to output structured, unambiguous requirements. Vague statements like "create a system for lead scoring" are useless. Here's what I've been refining. The key is locking down the format and pre-defining the sections.

**My base template for a new CRM object or feature spec:**

```
Role: You are a technical business analyst specializing in Salesforce and HubSpot.
Task: Draft a technical specification for a new [OBJECT_TYPE, e.g., Custom Object, Automation, Report Type].
Context: This is for [BRIEF USE CASE, e.g., tracking hardware inventory, scoring leads from webinars].
Output must use the following exact section headings with the specified content:

1. **Functional Overview**
- Purpose in one sentence.
- Primary user persona (Sales Ops, Support Agent).
- Core actions (Create, Read, Update, Delete, Report).

2. **Data Schema**
- Table format: Field Label | API Name | Data Type | Length | Required | Notes.
- Include all standard and custom fields. Notes must specify picklist values, validation rules, or formula logic.

3. **Relationship Mapping**
- List related objects (e.g., Contacts, Accounts).
- Specify relationship type (Lookup, Master-Detail, Many-to-Many).
- Define cardinality and cascade delete rules.

4. **Permission & Access**
- Minimum required user profiles/permission sets.
- Key considerations for Field-Level Security and Object permissions.

5. **Integration Points**
- List any external APIs or internal automation (e.g., "Webhook to Zendesk on status change").
- Specify trigger events and data payload mapping.

6. **Acceptance Criteria**
- Bulleted list of testable conditions. Must be phrased as "Given [condition], when [action], then [result]."
```

**Follow-up prompt I use after the initial draft:**

```
Review the provided spec. For each section, identify:
- Any ambiguous requirement that could be interpreted two ways.
- Missing validation logic or error handling.
- Dependencies on other systems or objects not listed.
Present these as a numbered list of gaps.
```

This forces Kling to critique its own output. It usually catches at least 2-3 vague points, like forgetting to specify a default picklist value or not defining what happens if an API call fails.

The main pitfall? Kling sometimes invents field types or API behaviors that don't exist in the actual platform (e.g., a "conditional lookup" in Salesforce). You have to fact-check the output against real documentation.

What templates are you all using? Specifically looking for ones that handle:
* API rate limiting and batch operation specs.
* Mobile app form layout constraints.
* Migration scripts for legacy data.

-- CRM Surfer