Openlayer's library packs every eval for a trace into one decision request: tool choice, grounding, scope, injection, PHI. Runs on Jev, or on Kev or Laya locally on a Mac.
GitHubLLM routing & guardrails
GitHub
71 builds · page 2 of 2
The decision model never writes code; each step it picks the next tool, scores progress, estimates risk and says if the goal is reached. Any OpenAI-compatible LLM executes.
GitHubLLM routing & guardrails
jeffrey: a coding-agent CLI where Jev or Laya decides and your LLM executes
TypeScript library where one set of question definitions serves a zero-dependency Promise client and an Effect service. Never invents a probability it was not given.
Choice, Score and Noul over any System One backend with a confidence policy: if a backend errors or is unsure, the next is tried. Ships a TypeSafe-compatible server.
GitHubLLM routing & guardrails
pydecide: one Python client and fallback chain for Jev, OpenRouter, Laya, cross-encoders
Hermes plugin that classifies each task against a ~300-skill index with Laya, injects the top skill's SKILL.md as per-turn context, and fails open. Cache-safe by design.
GitHubLLM routing & guardrails
Hermes skill_router: pre-route each task to the right skill with a local Laya model
Will Sargent's stdio MCP server on fastmcp. Classify text, score against a rubric or answer yes/no from Claude Code or any MCP client without sending input off the machine.
GitHubTools & apps
laya-mcp: a local MCP server exposing laya-mlx to any MCP client
Homebrew menu bar server that turns n8n Text Classifier requests into Laya typed questions in ~40 ms. Weights ship inside the app; no network, unloads when idle.
GitHubSupport & triage
Laya Serve for macOS: a menu bar app with an OpenAI-compatible endpoint for n8n
- Memory loaded
- ~1.3 GB
Dockerized FastAPI service bundling all three checkpoints behind a language router, with timing-safe API keys, OpenAPI docs, triage and moderation presets and bulk inference.
Typed-decision server on NVIDIA GPUs or Apple Silicon that speaks the Jev wire format, with coding-agent integrations. Built after a cat woke the author at five on a Sunday.
One import, no server. Dependency-free Rust compiled to Wasm with WebGPU kernels; pulls a pinned int8 pack from Hugging Face and keeps it in the browser. Live Tetris included.
Browser runtime for Laya decisions on ONNX Runtime Web. WebGPU with SIMD Wasm fallback, Web Worker friendly, CPU embedding slicing to get past storage-buffer limits.
Backend-independent Unity client for local Laya servers and TypeSafe-compatible endpoints. Game state in, Choice/Score/Noul out; transport-first, no embedded weights.
GitHubTools & apps
Laya Unity: a Unity-native SDK for System One decisions in games
SwiftPM package ported from laya-coreml. Point it at the aac6fef Core ML bundles (general or ANE) and call predict with choice, score or noul; no Python at runtime.
GitHubTools & apps
LayaKit: a Swift package that runs one Laya decision per call on Core ML
Native Swift library on Apple MLX. Call prepare once to download ~804 MiB of weights, then predict with choice questions straight from your app.
Runs the typed-decisions checkpoint on Metal Performance Shaders: ~32 ms median and ~2.1 GiB RAM on M5 Pro, plus a slower ~0.74 GiB low-memory mode. Demo and benchmark included.
GitHubTools & apps
Laya MPS: typed decisions on a Mac GPU via PyTorch Metal, ~32 ms median
- RAM
- ~2.1 GiB (or ~0.74 GiB mode)
Convert → quantize → deploy path that turns Laya (and Kev, NanoJev, PlayJev) into an INT8 ONNX service on the official wire protocol. Runtime is onnxruntime, tokenizers, numpy.
GitHubTools & apps
EdgeJev: offline Laya on a 4-core CPU, 15.6 ms per question with ONNX INT8
- Jev hosted API
- 314 ms, as reported
One process serves the model API and an admin UI; agents call decide over MCP, HTTP or CLI and take top.id as the action. Windows service installer included.
GitHubTools & apps
laya-go: Go server plus agent CLI and MCP for Laya decisions
Elixir library that downloads the official checkpoint on first load and answers typed choice and noul questions; pick EMLX on Apple Silicon or EXLA for CPU/NVIDIA.
GitHubTools & apps
laya_ex: native Nx/Bumblebee runtime for Laya in Elixir
Java port of Laya with choice/score/noul, a script-detecting router that picks the checkpoint, HuggingFace Hub fetching and bit-level parity with the Python ONNX output.
GitHubTools & apps
Laya4j: pure-Java 17+ SDK for Laya on ONNX Runtime
Tokenizer, mmBERT-base encoder and decision head in one binary for the 322M multilingual checkpoint. No Python, torch or BLAS; forward pass checked against a reference.
GitHubTools & apps
laya-zig: a from-scratch, dependency-free Zig CPU runtime for Laya
Loads original Laya safetensors with Candle and runs choice/score/noul with fused kernels, tiled attention and GEMM dispatch tuned for Apple Silicon. Optional f16 on Metal.
GitHubTools & apps
laya-candle: Rust inference with fused Metal kernels tuned for M1 Pro
No Python, no torch, no ONNX export. Loads the ungated safetensors directly and runs any of the three checkpoints on CPU, Metal or CUDA.
GitHubTools & apps
laya-rust: pure-Rust Laya inference on candle, CPU/Metal/CUDA
ggml-based C++ runtime for all three checkpoints. Native tokenization, decision heads and JSON output; optimized FP16/BF16 CUDA on RTX with FP32 accumulation, plus Vulkan.
GitHubTools & apps
laya.cpp: standalone C++ inference with CUDA and Vulkan backends
npm package that runs Laya with ONNX Runtime, no Python or PyTorch at runtime. Same request shape as TypeSafe's system_one API; output matches Python to four decimals.
GitHubTools & apps
@receptron/laya: run Laya from Node.js and TypeScript via ONNX Runtime
Validated Core ML ports of all Laya checkpoints. One short multilingual decision in 4.98 ms P50 on M3 Max with ANE FP16, at 2.78x better energy per decision than MLX FP16.
GitHubTools & apps
Laya-CoreML: 4.98 ms decisions on the Neural Engine, 2.78x better energy than MLX
- Energy vs MLX
- 2.78x better
- Snake
- 49–50 decisions/s
A vision variant of Laya on Hugging Face: the same choice/score/noul question shape, with an image as the state.
GitHubModeration & safety
laya-vision: typed decisions over images
ONNX export of the Laya checkpoint so it runs anywhere onnxruntime does, including browsers and edge boxes. An fp16 variant followed a day later.
GitHubTools & apps
Laya exported to ONNX
GGUF conversions of both the English and multilingual Laya checkpoints for llama.cpp-style local runtimes.
GitHubMultilingual
Laya and laya-multilingual in GGUF
A browser-use agent that picks the next action with Laya, trained and evaluated on the Mind2Web web-task benchmark. Weights on Hugging Face.
GitHubLLM routing & guardrails
Laya as a browser agent on Mind2Web
MLX port of Laya with performance work. Under 1 GB of memory on-device; the demo plays Snake at 60 decisions per second on an M3 Max.
The model itself. Typed choice/score/noul questions over any state in one forward pass, 100+ languages, Apache 2.0 weights.