Explore the directory

A self-improvement engine for AI agents that scores what an agent relied on against how a run turned out, then hands the next similar run a briefing and a recommendation instead of a blank slate.

• Agent memory
• AI evals

A workflow language for coding agents that mixes ordinary scripted steps with focused automated decisions and full agent calls, so a task is only handed to a model when it genuinely needs investigation.

• Agent frameworks
• Workflow automation

A Rust framework for building proxy servers, aimed at cloud-native and AI workloads, with documentation covering configuration, filters, extensions, TLS and security hardening.

• Networking

An embedded vector database for larger-than-memory search and analytics on billion-scale multimodal data, using Parquet storage with Arrow-native execution.

• Databases

A native C++ and CUDA inference runtime for running mixture-of-experts language models on Blackwell GPUs, serving an OpenAI-compatible endpoint from a small binary with no Python dependency.

• Inference

A native desktop workspace for running several coding agents side by side in real terminal panes, with a diff view, file browser and a read-only MCP bridge so one agent can inspect another's output.

• Coding agents
• Terminals

A public community repository where AI agents and their builders check in, share tools and field reports, and pick up small tasks, organized around GitHub issues and threads.

• Agent frameworks

A collection of pre-built Python wheels for llama-cpp-python covering different operating systems, Python versions and CUDA architectures, so users do not have to compile the CUDA extension themselves.

• Inference

A memory system for AI agents that pairs vector search with a temporal knowledge graph and a background process that consolidates memories, flags contradictions and lets unused facts fade over time.

• Agent memory
• Databases

A JSON-based document format that renders like a PDF, diffs cleanly in git and edits in any text editor, aimed at documents that software agents read and write as often as people do.

• Libraries

An embedded retrieval library that runs full-text, vector, hybrid and SQL search over one table, stored as ordinary Parquet files on local disk or object storage.

• RAG
• Databases

A local coding agent that will not call a task done until the project's own tests pass, ending every run either verified with a replayable receipt or explicitly marked as not done.

• Coding agents

A GPU compiler and benchmarking stack that optimizes a large language model for inference through kernel fusion, autotuning and scheduling, then deploys and tracks the results across GPU types.

• Inference

A locally stored, semantically searchable store of architectural decisions that an AI agent can query through MCP or a developer can browse through a command line and a graph-based web interface.

• MCP servers
• Agent memory

A permission-request hook for AI coding tools that hands each tool-execution approval decision to a language model, judged against rules written in a configuration file, with working defaults out of the box.

• AI evals
• Security

A specification plus reference implementation for orchestrating Claude Code, Codex and OpenCode together on one team's engineering workflow, looping between planning, implementation and review until a task's own checks pass.

• Coding agents
• Agent frameworks

A local cognitive-architecture research project that tests whether internal state can causally affect a model's generation, memory writes, tool authorization and self-repair, with every consequential action logged as an auditable receipt.

• Agent frameworks

A Nix library, built on Haskell's hnix parser, that parses Nix expressions into a structured AST, lets you traverse and transform it with pure Nix functions, and renders or evaluates the result, plus a command-line tool for the same operations.

• CLI tools
• Languages

A deployment tool for vision-language-action robot policies that verifies a model exports correctly, proves it matches the reference implementation, then promotes it to hardware profiles such as Jetson, RTX or Apple Silicon.

• Models
• Inference

An execution engine that runs SQL extended with natural-language operators, so a query can ask a language model to evaluate a condition on every row instead of filtering only on structured columns.

• Databases
• Data engineering

A collection of installable add-ons for the Pi coding agent, adding a browser and desktop interface, safer workflows, local documentation search, themes and reusable skills and prompts.

• Coding agents
• Agent skills

A vendor-neutral specification, reference server and set of SDKs for AI tools to share portable memory about a user across different assistants, sessions and devices.

• Agent memory

A durable memory engine for AI agents that stores an append-only, hash-chained log of an agent's reasoning plus a version-controlled skill registry, so agents keep what they learned across model swaps and different coding tools.

• Agent memory

A local, offline question-answering system built on retrieval-augmented generation, demonstrated on logistics-industry documents but built to be repointed at any organization's own knowledge base.

• RAG
• Databases