AI tool

Open-source test runner for AI agent skills and multi-step workflows. It runs the same scenario across supported harnesses, captures evidence from tools, commands, files, and transcripts, and lets you assert on the observed behavior.

Browser-local tool that checks AI agent instructions for missing written guardrails against 23 public controls. It supports pasted prompts, local files, and public GitHub files, with CLI and GitHub Action options for recurring checks.

Parley is coordination infrastructure for autonomous coding agents. It lets agents hand off work asynchronously, records each relay in an append-only flight recorder, supports escalation, soft file-locking, and works with Claude Code, Cursor, Codex, Copilot, Antigravity, and any MCP client.

Ambient macOS light strip for AI coding agents. Shows agent state beside the Dock with eight virtual LEDs and supports publishing status over a local CLI/API.

unYOLO is a framework for building credential brokers and proxies for AI agents. It keeps provider credentials out of the agent, applies local policy rules, supports timed grants and operator approvals, and includes brokers for GitHub, Hugging Face, and approved Unix commands.

An open 30B agentic model from Meta designed to run locally on consumer hardware for coding, function calling, and other always-on agent workflows.

Pier gives each coding-agent session its own cloud VM on the user’s AWS account, with parking, resuming, and repo setup built in.

Stonefold is a deterministic enforcement gateway for AI agents. It sits between an agent and real systems, validates typed intents, applies policy, supports human approval and kill-switch controls, and records an audit trail for every attempt.

An autonomous audit-fix-verify loop for Claude Code that runs iterative code improvements with verification gates and a public journal of runs.

Doubt is an open-source, zero-dependency CLI and Agent Skill for turning contested questions into source-grounded evidence maps with supports, contradictions, unknowns, and portable HTML output.

A lightweight API key proxy for self-hosted llama.cpp servers with per-user token usage tracking, rate limiting, and multi-server routing.

A local-first TypeScript toolkit for crash-testing AI agents and MCP servers, replaying failures deterministically, and turning broken traces into regression tests.

Engram is an open-source learning engine for Claude Code and other agentic coding platforms. It turns AI help into a study loop with tutor-style prompts, blind recall checks, and spaced-review scheduling so concepts stick longer.

A local merge queue for parallel Claude Code agents that serializes landings, builds, and test runs to avoid push races and shared-resource flakiness.

A private, local-first AI voice journal that turns notes into structured memory you can query later. It focuses on grounded recall across your own entries, with voice capture, semantic search, export controls, and mobile apps for iPhone and Android.

Headroom is a macOS menu bar app that compresses noisy prompt context before it reaches Claude Code and Codex, aiming to cut token usage by about 50% while keeping the original retrievable on demand.

Forkbench is a native Mac control room for running multiple coding agents in parallel, each in its own git branch, from a single board. It adds shared coordination, notes, secrets handling, worktree controls, and remote oversight for supervising agent-heavy workflows.

CMEM is an AI agent memory layer for coding agents, with an open-source local engine and a managed cloud service that syncs observations through a private MCP link. It supports Claude Code, Cursor, Windsurf, Codex, Gemini CLI, and other MCP clients.

Desktop app for running CLI coding agents like Claude Code and Codex in parallel git worktrees, with built-in diff review, multi-repo support, broadcast, and sandboxing.

cloro is an API for scraping and structuring results from search and AI search surfaces like ChatGPT, Gemini, Perplexity, Copilot, Google Search, and AI Overview. The page shows request and response examples, supported models, pricing tiers, and claims of parsed markdown, sources, and citations.

Open-source, self-hosted MCP knowledge server for AI agents that adds read-only access to company data, preserves business context, and can publish live dashboards as apps.

A public MCP server from PixelGust that lets Claude, Cursor, and other MCP clients query live weather, forecast, climate history, terrain, fire risk, flood susceptibility, and related environmental data.

Open-source causal-graph memory for AI agents with MCP support, scoring, active forgetting, and benchmark results published in the repo.

A 30-second WebGPU benchmark that measures a device’s memory-bandwidth ceiling for local AI and checks for a WebGPU subgroup correctness bug, with receipts, projections, and comparison data.