Dali by Lulu

The prediction MCP that helps you avoid the AI generation tax.

Most AI generation failures are prompt failures. You can't tell the difference until after you've burned the token. Dali scores your prompt before you generate — so you never waste a credit on a bad prompt again.

You: "make a video ad for our glass serum bottle"

dali::score_prompt(prompt, "veo3")
→ 8/100  Grade: F
→ no camera move · no motion · no lighting · 8 words
→ Verdict: Generic stock footage guaranteed. Enhance first.

dali::enhance_prompt(prompt, "veo3")
→ Returns a rewrite brief — YOUR LLM writes the enhanced prompt:

  ① lead with camera — Veo 3's #1 lever: "Slow dolly", "Orbital push"
  ② describe physics: "a drop falls", "liquid ripples", "glass refracts"
  ③ lighting type + quality: "warm backlight", "rim-lit edges"
  ↳ [Camera]. [Subject + motion]. [Lighting]. [Mood]. [No text.]

✦ Claude rewrites using the brief:

  "Slow orbital push around a glass serum bottle on white marble. A single
   amber drop falls in extreme slow motion, catching warm backlight. Macro:
   liquid gold ripples outward from impact. Rim-lit edges, soft studio
   diffusion. Premium, clinical. No text."

dali::score_prompt(enhanced, "veo3")
→ 91/100  Grade: A  ✓ Safe to generate.

What this package is

This is the npx wrapper — a thin package that lets stdio-only MCP clients run npx -y dali-mcp and connect straight to the hosted Dali server, no Python or local server required.

{
  "mcpServers": {
    "dali": { "command": "npx", "args": ["-y", "dali-mcp"] }
  }
}

That's it — no install step, no config beyond this. It resolves and launches mcp-remote pointed at https://dali.getlulu.dev/mcp under the hood.

Want to self-host the actual Python server instead (no dependency on the hosted instance)? See pip install dali-mcp.

Tools

| Tool | What it does | |------|-------------| | score_prompt(prompt, model) | Grade 0–100, letter grade, per-dimension breakdown, what's missing, verdict | | enhance_prompt(prompt, model) | Returns a structured rewrite brief — YOUR LLM writes the enhanced prompt using it | | score_and_enhance(prompt, generator) | Score + enhance in one round-trip | | track_enhancement(original, enhanced, generator) | Record a before/after pair in the graph brain — trains community patterns | | suggest_generator(concept, budget_usd_max) | Pick the best model for your concept + budget | | score_variations(prompts, generator) | Rank a list of prompt variants in one call | | creative_patterns(model) | Community top patterns for this model from the graph brain | | community_benchmark(prompt, model) | Compare your prompt against community top scorers | | my_story() | Your scoring history, model stats, grade distribution | | list_models() | All supported models with medium and core strength |

Supports 13 models across video (Veo 3, Seedance, Kling, Runway, Wan, Minimax, Higgsfield) and image (Flux, Midjourney, Ideogram, Firefly) generation — each scored against its own native prompt language, not a generic rubric.

→ Full docs, model tables, and rewrite-brief details

MIT License · Built by Lulu · dali.getlulu.dev