Add Mind programming language

[star-ga]

Language Overview

Mind (Machine Intelligence Native Design) is a systems programming language designed for AI/ML and numerical computing. It features:

• Static tensor types with compile-time shape checking
• Reverse-mode automatic differentiation
• MLIR/LLVM backend for optimized code generation
• GPU device placement with on(gpu) blocks

Public Repositories Using Mind

Repository	Description	.mind files
cputer/mind	Language compiler and runtime	26+ files
cputer/mind-spec	Language specification	8 files
cputer/nikolachess	NNUE chess engine	15 files
cputer/mind-ray	GPU path tracer	12 files
cputer/fractal-voyager	WebGPU fractal explorer	5 files

Total: 60+ .mind files across public repositories

Note: The Mind compiler and runtime are about to be released publicly. Additional private repositories contain substantial Mind codebases used in production. We're adding language support ahead of the public release to ensure GitHub syntax highlighting is ready for users.

Language Definition

```yaml
Mind:
type: programming
color: "#7b2ff7"
extensions:
- ".mind"
tm_scope: source.mind
ace_mode: text
```

Files Included

• `lib/linguist/languages.yml` - Language definition
• `vendor/grammars/mind-grammar` - TextMate grammar submodule
• `samples/Mind/` - 10 sample files, 3,490 lines of production code:

Sample Files

File	Lines	Source	Description
`search.mind`	829	nikolachess	Draw-first alpha-beta search with TT
`tensor_board.mind`	540	nikolachess	GPU-accelerated 16-channel board representation
`fractal.mind`	418	fractal-voyager	Vectorized Mandelbrot/Julia/Burning Ship
`material.mind`	343	mind-ray	Physically-based materials (diffuse, metal, glass)
`render.mind`	333	mind-ray	Monte Carlo path tracing with Russian roulette
`vec3.mind`	290	mind-ray	3D vector math library
`camera.mind`	247	mind-ray	Pinhole camera with look-at positioning
`complex.mind`	190	fractal-voyager	Complex number tensor operations
`conv2d.mind`	171	mind-spec	2D convolution with stride/padding
`mlp_backward.mind`	129	mind-spec	MLP backpropagation with autodiff

Syntax Highlights

• Rust-like syntax with `fn`, `let`, `struct`, `enum`, `match`
• Tensor type annotations: `Tensor<f32, [2, 3]>`, `tensor<f16, (8, 8)>`
• Generic shape parameters: `Tensor<f32, [batch, hidden_dim]>`
• GPU placement: `on(gpu0) { ... }`
• CUDA kernels: `@cuda_kernel fn name(...) { ... }`
• Automatic differentiation: `grad()`, `backward()`
• Module imports: `import std.tensor;`

Links

• Website: https://mindlang.dev
• Grammar repo: https://github.com/cputer/mind-grammar
• License: Apache 2.0 (compiler & language); commercial terms for enterprise extensions

[lildude]

This PR is invalid. You have not followed the instructions in the CONTRIBUTING.md file and then have chosen to ignore our PR template.

Your referenced language repo is less than a month old which means this is a very new/hobby language. As detailed in the CONTRIBUTING.md file, we have minimum usage requirements. This means we do not accept PRs for very new or hobby languages. Closing but feel free to open a PR when your language is popular enough, and make sure you follow the instructions an use our PR template.

[star-ga]

Thanks for the review and explanation. I understand - we jumped the gun on this one.

We'll circle back once Mind has more community adoption and meets the usage requirements. Appreciate you taking the time to explain the process.