llm

Swift package for building provider-agnostic AI agents on Apple platforms, with tools for runtime permissions, hooks, context compaction, session state, validation loops, and traceable execution.

TypeScript middleware that trims or compresses LLM chat history to fit a token budget, with support for OpenAI, Anthropic, and AI SDK integrations.

Open-sourced local multi-agent OS that runs three agents on consumer hardware, lets them form goals from environmental pressure, and supports self-modifying workflows through tool creation and change requests.

A research-style demo showing LLM agents evolving through a genetic algorithm to play Pokémon in parallel sandboxed VMs. The page explains the setup, scoring loop, generations, results, and links to the repo and run artifacts.

A ~1000-line C workflow runtime where an LLM can modify a DAG during execution using a restricted four-verb grammar, with every change written to an append-only audit log for replay and inspection.

Open-source Rust coding agent with LLM-native code understanding, shell safety, and support for multiple LLM providers with automatic failover.

Lightport is a lightweight AI gateway that makes LLM providers OpenAI-compatible. The repo shows supported endpoints, a long provider list, quickstart commands, request examples, and active maintenance signals.

Axminster is an AI-augmented software design tool for model-driven engineering. It supports UML 2.5.1 and OCL 2.4, can save models as XMI, and includes an MCP server for sharing designs with LLM-based coding assistants.

A blind listening arena for AI-composed music. Users compare two pieces against the same brief, vote A/B/tie/neither, and then reveal results and rankings.

Uno is a research repository for speeding up LLM inference with discrete diffusion and lossless multi-token decoding. The repo includes inference, training, and evaluation code, plus installation steps, checkpoints, and example workflows.

Local privacy gateway that redacts sensitive text before sending prompts to external LLMs, then reconstructs the final answer locally.

Dependency-free Python library that compiles a declarative JSON query format into safe, parameterized SQL for LLM workflows, REST APIs, and frontend-backed data tables.

ContextSpy is an LLM context profiler that intercepts agent traffic, breaks down token usage, and visualizes how prompts change across requests and sessions.

An AI-powered Obsidian plugin that turns notes into a self-organizing wiki with auto-generated entity pages, bidirectional links, traceable answers, and local-first support.

A governance plugin for OpenClaw that registers instances with a Recursant registry and intercepts tool calls, LLM prompts, and chat messages for auth, PII redaction, rate limiting, and auditing.

Single-binary viewer for HTML reports generated by AI agents, with local browsing, live previews, search, and shareable links.

A hands-on workshop repository that walks developers through building a small GPT-style language model from scratch, including tokenization, transformer blocks, training, and text generation.

Browser-based RAG app that runs entirely client-side, using BM25/Lunr.js for retrieval and direct LLM provider APIs for answering questions without a backend.

A GitHub repository for a 7-node ESP32-S3 cluster that runs a small LLM using 1.58-bit BitNet-style quantization over an SPI daisy-chain.

A text-first, hands-on guide to agentic AI for practicing engineers, with runnable local-first code samples and a module-based structure covering LLM basics, tool calling, RAG, memory, multi-agent patterns, evaluation, safety, and enterprise topics.

A step-by-step tutorial that walks readers through making a first Anthropic Claude API call in Python, including environment setup, client creation, token usage, and printing the response.

Open-source LLM uncertainty and hallucination detection tool that estimates epistemic risk in microseconds using exact-match normalized entropy, with Rust gateway and Python package support.

vernLLM is a TypeScript LLM call framework that keeps retries, fallback, rate limiting, circuit breaking, caching, and middleware inside your own process. It supports OpenAI-compatible providers plus Anthropic, Gemini, Bedrock, and others via adapters.

Live demo of a prompt-injection defense for LLMs using per-position semantic overlays. The page shows how marked spans can be treated as untrusted while the base model stays frozen, with interactive examples for retrieved text, hidden instructions, language marking, and request-scoped overlays.