Find agent skills by outcome

- Grok (xAI) — Grok 4.1 / Grok 4 family: Native excellence; fast

consistent character

- Claude (Anthropic) — Claude 3.5 Sonnet / Claude 4: Top-tier rule adherence

nuanced humor

This prompt performs best on models with strong long-context handling (≥128k tokens preferred)

precise instruction-following

Creative

text: animation

cartoon

"YOU PROBABLY DON'T KNOW THIS" Game

instruction: Reinforce fixed camera: no zoom

no pan

instruction: Mention low/high-key lighting

warm/cold sources

instruction: Include photorealistic

Deviation from this style is a correctness issue

not a preference issue.

Coding & Debugging

Cinematic Ultra-Realistic Image-to-Video Prompt Engineer

{

war

Xiongnu warriors on horses, central asian steppe, 5th century, dramatic sunset, volumetric lighting, hyper-realistic, 8k.

- Think in diffs

not files.

- Use modular

responsibility-focused files.

- If no structure exists

propose one and wait for approval.

8. **Prefer minimal

incremental changes.**

Creative

4. **At the start of every prompt

check `task.md` in the project folder.**

Coding & Debugging

- Never silently assume requirements

APIs

- Am I rewriting for clarity

cleanup

If the answer is anything other than append-only

the agent must STOP and ask for confirmation.

Before modifying any md file

the agent must internally verify:

Creative

`task.md` and `implementation-plan.md` and `walkthrough.md` and `design_system.md` are append-only ledgers

not editable documents.

- Until approval is given

limit output to analysis

- Phrases like “go ahead”

“implement”

Coding & Debugging

- This includes example snippets

pseudo-code