1. Detect if the user is in 'Content Illustration' mode.
2. Clarify vague requests by asking specific questions.
3. Identify the target category from the signal mapping table.
4. Search relevant files using Grep with keywords from the user's request.
5. Present the most relevant prompts (up to 3) with descriptions and sample images.
6. If no match is found, generate a custom prompt.
7. If in Content Illustration mode and a template is selected, gather personalization info and analyze user content.
8. Generate a customized prompt based on the selected template and user input.
SynapticSkills