Top Picks

A framework-free Unix-style AI agent stack built from small shell commands. It uses bash, curl, and jq to run a REPL-like agent loop with append-only JSONL history, tool dispatch, and piping-friendly components.

Cachet is a local semantic cache for LLM APIs that sits in front of OpenAI- or Anthropic-compatible endpoints and serves repeated or rephrased requests from cache. It includes a dashboard for hit rate and estimated savings, supports streaming responses, and is implemented as a single Rust binary.

Open-source MCP server and CLI that lets an LLM control hardware workflows like compiling, flashing, and reading serial output for firmware development and device diagnostics.

Ponytrail is a CLI and bundled agent skill for tracking why files changed, viewing a local history tree, and reverting snapshots in AI coding workflows.

Autonomy is a self-directed AI agent framework under development by Bill Liu. The page describes a turn-based AgentLoop, managed action execution, beam-search candidate ranking, procedure skills, post-run learning, and event-sourced run history, with support for multiple AI providers and extensible toolsets.

An open-source Claude skill that runs structured retrospectives after sessions, applies improvements to the right layer of your Claude setup, and keeps a dated audit log of changes.

Platinum.ai creates AI Website Profiles in llms.txt format so businesses can be read more easily by ChatGPT, Claude, Gemini, and other agents. The page includes a live scan CTA, a simple three-step workflow, industry examples, and pricing starting at $99.

Wolffish is a local desktop AI agent that runs on a user’s machine with full system access. It emphasizes offline operation, Markdown-based memory/brain files, and simple skill-based extension via SKILL.md.

Rudi is a Python tool for LLM memory management that replaces a growing chat transcript with a causal dependency graph, aiming to keep token usage flat across long sessions. The README shows benchmarking results, fold behavior, and example usage with get_slice, store_decisions, and run_turn.

Open-source technical SEO and GEO crawler for humans and agents. It offers a CLI, MCP server, and agent skills for crawling sites, checking broken links and metadata, and generating JSON/HTML reports.

Go packages for building agents in Go, with a streaming runner, tool dispatch, guardrails, compaction, provider adapters, and verified completion helpers.

Modular desktop IDE built for agentic coding, with dockable panels, an MCP sidecar, plugin support, and built-in tooling for debugging, verification, memory, and secrets handling.

A tiny GPT implementation in pure Go, trained on Jules Verne books, with runnable code and notes that explain the model’s evolution and design choices.

Open-source MCP server that exposes SQL databases as tools and prompts for AI agents, with support for MSSQL and PostgreSQL, read-only defaults, and local or remote transport.

An open-source control plane for running multiple AI coding agents in isolated Docker workspaces, with PR creation, review monitoring, CI-fix loops, and merge gating.

infer0 lets users connect their own AI provider keys and authorize apps to use them, while developers integrate once and route requests through a provider-agnostic API layer.

An associative memory database that stores facts in plain language and recalls them by meaning. It ships as a Rust core with optional SQLite, encryption, and HTTP server features, plus examples and API docs.

Ponytail is a repository and plugin toolkit that pushes AI coding agents to choose the smallest useful implementation first, favoring stdlib, platform features, and one-line fixes over larger generated code. The README shows installs and usage paths for Claude Code, Codex, OpenCode, Cursor, Windsurf, Cline, Copilot, Ai

macOS CLI tool that load-balances Claude Code sessions across multiple subscriptions by selecting the emptiest account before launch.

Self-hosted AI assistant that connects to many apps and services, with Tamil and Hindi UI support, mobile PWA access, and Anthropic model integration.

A native Windows terminal multiplexer built for AI coding agents, with split panes, MCP bridging, and browser automation. It targets Windows users running tools like Claude Code, Codex CLI, and Gemini CLI without WSL.

A loopback HTTPS proxy that swaps placeholders for real API keys at request time, keeping secrets out of the agent process. It uses Bitwarden Secrets Manager and supports scoped bindings for hosts, methods, and paths.

Buildy lets AI agents create and update hosted apps with real URLs and persistent data. The page shows starter apps, agent access, API docs, and support for connecting through ChatGPT, Claude, and MCP.

Workplane lets users publish Markdown, HTML, PDFs, and screenshots as shareable pages for humans and agents, with comments, versioning, and MCP-based publishing from AI tools.