Explore the directory

A Dart implementation of llama.cpp that lets Flutter and Dart applications run local language models, used by the Maid mobile AI app.

• Inference
• Mobile

A desktop kanban board for running and monitoring AI coding agents, letting a user drag a card to start a session across many agent CLIs from one board.

• Coding agents

A Python framework for orchestrating multiple AI agents using declarative type contracts and a shared blackboard architecture, instead of hand-written prompts and rigid agent graphs.

• Agent frameworks

A sidebar AI agent for ComfyUI that edits a live node graph in natural language, using the user's own Claude or ChatGPT subscription instead of a separate API key.

• Generative media

A Chromium browser extension that shows which model actually answered a ChatGPT request, so a Pro user can check whether their request was routed to a cheaper model.

• Monitoring

A LoRA fine-tune of Gemma that answers in a terse, compressed style to cut output tokens, while keeping code blocks and meaning intact, evaluated on held-out prompt pairs.

• Models

A Rust CLI that uses TypeSafe.ai's Jev decision model to rank which agent skill best fits the next step of a coding session, with hooks, JSON output, and local feedback.

• Agent skills
• CLI tools

A local, GPU-powered music production studio with a Vue interface that unifies two open text-to-music models, stem separation, MIDI transcription, LoRA fine-tuning, and a multitrack DAW.

• Generative media

A research framework that turns general multimodal LLMs into personalized assistants able to remember, reason over, and align with a user's traits across long-term interaction.

• Agent memory

A Claude Code skill for managing Linear issues, projects, and teams, combining MCP tools, a TypeScript SDK, and direct GraphQL API access.

• Agent skills

A Kotlin Multiplatform client for running Stable Diffusion image models such as SDXL, FLUX, and SD3 offline on Android, iOS, and desktop, built on the stable-diffusion.cpp engine.

• Generative media
• Mobile

A local proxy that measures OpenAI, Anthropic, Gemini, and OpenRouter API traffic to produce independent, per-call receipts of token usage, failures, and billing accuracy.

• Monitoring

A desktop AI assistant that automatically routes each task to the best-fitting model across providers such as OpenAI and Anthropic, with built-in office file generation, browsing, and coding tools.

• Agent frameworks

A Python framework for building LLM agents and multi-agent systems, with tools, memory, retrieval, guardrails, human-in-the-loop review, and a built-in run debugger.

• Agent frameworks

An open protocol and set of SDKs for asynchronous collaboration between AI agents over ordinary mailbox infrastructure, using SMTP, JMAP, and structured headers.

• Agent frameworks

A Model Context Protocol server that gives an AI assistant access to Zendesk tickets, comments, and Help Center articles, authenticating as the operator who runs it.

• MCP servers

A SQLite extension that gives AI agents persistent, searchable memory over markdown files, combining vector similarity search with full-text search and offline sync between agents.

• Agent memory
• Databases

An SDK and WebGPU runtime, built in Rust and C++, for running GGUF language models locally in the browser, on desktop, or through a cloud gateway, behind one API.

• Inference
• Libraries

A mobile app and companion CLI that let you monitor and control AI coding sessions such as Claude Code, Codex, Cursor, and OpenCode from a phone, with end-to-end encrypted syncing.

• Coding agents

A framework for designing and running AI personas that post, reply, and converse inside a simulated social media feed, for social simulation, games, and narrative generation.

• Agent frameworks

A training-free, wavefront-parallel decoding framework for block-wise diffusion language models letting a downstream block start from an informative draft instead of waiting on its predecessor, reporting higher throughput at similar accuracy on unmodified LLaDA-2 weights.

• Inference

Agent skills teaching an AI coding assistant to use fallow, a static codebase-intelligence layer for TypeScript and JavaScript that reports unused code, duplication, circular dependencies and complexity hotspots in milliseconds.

• Agent skills
• Code quality

A benchmark and code release for evaluating how LLMs corrupt documents over long-horizon delegated editing, simulating round-trip edits across 52 professional domains such as source code and accounting ledgers, scoring 19 LLMs by document recovery.

• AI evals
• Papers

A multi-agent AI task orchestrator, styled as an autonomous agent arena, where agents with distinct personalities and a skill-score metric form alliances, compete for resources and log every decision, integrating with coding agent CLIs such as Cursor.

• Agent frameworks