[IMPORTANT] Use TaskCreate to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI may ask user whether to skip.
Quick Summary
Goal: Write and optimize prompts for AI text, image, and video generation models (Claude, GPT, Midjourney, DALL-E, Stable Diffusion, Flux, Veo).
Workflow:
- Identify — Determine model type (LLM, image, video) and desired outcome
- Structure — Apply model-specific prompt patterns (Role/Context/Task for LLMs, Subject/Style/Composition for images)
- Refine — Iterate with A/B testing, style keywords, negative prompts
Key Rules:
- Use clarity, context, structure, and iteration as core principles
- Apply model-specific syntax (Midjourney
--ar, SD weighted tokens, etc.)
- Load reference files for detailed guidance per domain (marketing, code, writing, data)
AI Artist - Prompt Engineering
Craft effective prompts for AI text and image generation models.
Core Principles
- Clarity - Be specific, avoid ambiguity
- Context - Set scene, role, constraints upfront
- Structure - Use consistent formatting (markdown, XML tags, delimiters)
- Iteration - Refine based on outputs, A/B test variations
Quick Patterns
LLM Prompts (Claude/GPT/Gemini)
[Role] You are a {expert type} specializing in {domain}.
[Context] {Background information and constraints}
[Task] {Specific action to perform}
[Format] {Output structure - JSON, markdown, list, etc.}
[Examples] {1-3 few-shot examples if needed}
Image Generation (Midjourney/DALL-E/Stable Diffusion)
[Subject] {main subject with details}
[Style] {artistic style, medium, artist reference}
[Composition] {framing, angle, lighting}
[Quality] {resolution modifiers, rendering quality}
[Negative] {what to avoid - only if supported}
Example: Portrait of a cyberpunk hacker, neon lighting, cinematic composition, detailed face, 8k, artstation quality --ar 16:9 --style raw
References
Load for detailed guidance:
| Topic |
File |
Description |
| LLM |
references/llm-prompting.md |
System prompts, few-shot, CoT, output formatting |
| Image |
references/image-prompting.md |
Style keywords, model syntax, negative prompts |
| Nano Banana |
references/nano-banana.md |
Gemini image prompting, narrative style, multi-image input |
| Advanced |
references/advanced-techniques.md |
Meta-prompting, chaining, A/B testing |
| Domain Index |
references/domain-patterns.md |
Universal pattern, links to domain files |
| Marketing |
references/domain-marketing.md |
Headlines, product copy, emails, ads |
| Code |
references/domain-code.md |
Functions, review, refactoring, debugging |
| Writing |
references/domain-writing.md |
Stories, characters, dialogue, editing |
| Data |
references/domain-data.md |
Extraction, analysis, comparison |
Model-Specific Tips
| Model |
Key Syntax |
| Midjourney |
--ar, --style, --chaos, --weird, --v 6.1 |
| DALL-E 3 |
Natural language, no parameters, HD quality option |
| Stable Diffusion |
Weighted tokens (word:1.2), LoRA, negative prompt |
| Flux |
Natural prompts, style mixing, --guidance |
| Imagen/Veo |
Descriptive text, aspect ratio, style references |
Anti-Patterns
- Vague instructions ("make it better")
- Conflicting constraints
- Missing context for domain tasks
- Over-prompting with redundant details
- Ignoring model-specific strengths/limits
IMPORTANT Task Planning Notes (MUST FOLLOW)
- Always plan and break work into many small todo tasks
- Always add a final review todo task to verify work quality and identify fixes/enhancements
1---2name: ai-artist-33description: [AI & Tools] Write and optimize prompts for AI-generated outcomes across text and image models. Use when crafting prompts for LLMs (Claude, GPT, Gemini), image generators (Midjourney, DALL-E, Stable Diffusion, Imagen, Flux), or video generators (Veo, Runway). Covers prompt structure, style keywords, negative prompts, chain-of-thought, few-shot examples, iterative refinement, and domain-specific patterns for marketing, code, and creative writing.4license: MIT5---6
7> **[IMPORTANT]** Use `TaskCreate` to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI may ask user whether to skip.
8
9## Quick Summary
10
11**Goal:** Write and optimize prompts for AI text, image, and video generation models (Claude, GPT, Midjourney, DALL-E, Stable Diffusion, Flux, Veo).
12
13**Workflow:**
14
151. **Identify** — Determine model type (LLM, image, video) and desired outcome
162. **Structure** — Apply model-specific prompt patterns (Role/Context/Task for LLMs, Subject/Style/Composition for images)
173. **Refine** — Iterate with A/B testing, style keywords, negative prompts
18
19**Key Rules:**
20
21- Use clarity, context, structure, and iteration as core principles
22- Apply model-specific syntax (Midjourney `--ar`, SD weighted tokens, etc.)
23- Load reference files for detailed guidance per domain (marketing, code, writing, data)
24
25# AI Artist - Prompt Engineering
26
27Craft effective prompts for AI text and image generation models.
28
29## Core Principles
30
311. **Clarity** - Be specific, avoid ambiguity
322. **Context** - Set scene, role, constraints upfront
333. **Structure** - Use consistent formatting (markdown, XML tags, delimiters)
344. **Iteration** - Refine based on outputs, A/B test variations
35
36## Quick Patterns
37
38### LLM Prompts (Claude/GPT/Gemini)
39
40```
41[Role] You are a {expert type} specializing in {domain}.
42[Context] {Background information and constraints}
43[Task] {Specific action to perform}
44[Format] {Output structure - JSON, markdown, list, etc.}
45[Examples] {1-3 few-shot examples if needed}
46```
47
48### Image Generation (Midjourney/DALL-E/Stable Diffusion)
49
50```
51[Subject] {main subject with details}
52[Style] {artistic style, medium, artist reference}
53[Composition] {framing, angle, lighting}
54[Quality] {resolution modifiers, rendering quality}
55[Negative] {what to avoid - only if supported}
56```
57
58**Example**: `Portrait of a cyberpunk hacker, neon lighting, cinematic composition, detailed face, 8k, artstation quality --ar 16:9 --style raw`
59
60## References
61
62Load for detailed guidance:
63
64| Topic | File | Description |
65| ------------ | ----------------------------------- | ---------------------------------------------------------- |
66| LLM | `references/llm-prompting.md` | System prompts, few-shot, CoT, output formatting |
67| Image | `references/image-prompting.md` | Style keywords, model syntax, negative prompts |
68| Nano Banana | `references/nano-banana.md` | Gemini image prompting, narrative style, multi-image input |
69| Advanced | `references/advanced-techniques.md` | Meta-prompting, chaining, A/B testing |
70| Domain Index | `references/domain-patterns.md` | Universal pattern, links to domain files |
71| Marketing | `references/domain-marketing.md` | Headlines, product copy, emails, ads |
72| Code | `references/domain-code.md` | Functions, review, refactoring, debugging |
73| Writing | `references/domain-writing.md` | Stories, characters, dialogue, editing |
74| Data | `references/domain-data.md` | Extraction, analysis, comparison |
75
76## Model-Specific Tips
77
78| Model | Key Syntax |
79| ---------------- | --------------------------------------------------- |
80| Midjourney | `--ar`, `--style`, `--chaos`, `--weird`, `--v 6.1` |
81| DALL-E 3 | Natural language, no parameters, HD quality option |
82| Stable Diffusion | Weighted tokens `(word:1.2)`, LoRA, negative prompt |
83| Flux | Natural prompts, style mixing, `--guidance` |
84| Imagen/Veo | Descriptive text, aspect ratio, style references |
85
86## Anti-Patterns
87
88- Vague instructions ("make it better")
89- Conflicting constraints
90- Missing context for domain tasks
91- Over-prompting with redundant details
92- Ignoring model-specific strengths/limits
93
94---
95
96**IMPORTANT Task Planning Notes (MUST FOLLOW)**
97
98- Always plan and break work into many small todo tasks
99- Always add a final review todo task to verify work quality and identify fixes/enhancements
majiayu000