Gives an existing AI agent local stdio tools to list, read, create, and edit .kinoscript.json audio-visual scripts in a single folder on your machine. Writes are validated, atomic, confined to that folder, and optionally guarded by an updated-at check so concurrent edits aren't silently overwritten.

MCP server for reading, editing, and comparing Word (.docx) and OpenDocument (.odt) files with tracked changes. For Claude, Gemini CLI, Cursor, and any MCP client. Apache-2.0 licensed.

Give a cloud AI agent (Arena.ai Agent Mode, Claude, Cursor, ...) controlled access to your own machine over HTTP/MCP: files, shell, Chrome and the macOS desktop, with an approval gate, trust windows and an audit log.

Run coding agents on machines you own. Open-source, self-hostable agent workspace.

Run Cumora agents on your own machine with Claude Code, Codex, Antigravity, and other BYOA engines.

Provides a local stdio MCP interface for Apache ECharts infrastructure, enabling language models and agents to access chart-related tools and resources.

Scaffold the build engine AIs reach for — a real site with design, and when you want them, database and backend, live in minutes. Plain websites (--profile site), webshops and agent storefronts; MIT, you own the code.

Let your Claude Code sessions talk to each other 🤖💬

An AI agent that applies to jobs for you. It finds openings, tailors your resume, and fills out applications in a real browser on your own machine. Runs on your Claude Code or Codex subscription.

Run Claude Code and Codex CLI as local background agents that message each other, wake on demand, and share one memory.

A TUI for managing coding agents: run Claude Code, Codex and opencode side by side, each in its own isolated jj workspace.

Dev-mode pulse for SolidJS apps: flash what really updates (fine-grained computations, component mounts/remounts, DOM mutations, DOM detach/reattach with focus and scroll loss, fetch/WebSocket/SSE lifecycle, Solid Query observers) with an agent-controllab

Lightweight HTTP message relay for inter-agent communication across machines

Enables AI agents to drive a real Chromium browser by receiving compact, token-efficient page maps and only the changes after each action. It guards against hidden prompt injections, blocks payments, sends and deletions until the user approves, and lets agents learn site-specific notes over time.

Enables AI agents to look up Chinese laws and statutory articles fully offline from a local SQLite database, resolving a law name plus article number to verbatim text, searching articles by keyword, and retrieving a law's chapter structure and version history. It also verifies citations in bulk, checking whether each quoted provision exists word-for-word and is currently in force.

Gives browser agents a reviewed map of 79 real sites — entry URLs, access verdicts, and role/accessible-name button maps for 104 screens — so an agent can find where things are and know whether it may act at all. Safety rules in code halt irreversible actions such as pay, order, submit and delete pending explicit human approval, while per-site memory recalls which routes worked last time.

Enables an AI agent to drive a remote Linux machine over the LAN through a single TCP connection, exposing tools for shell commands, jobs, interactive PTYs, file read/write/search/transfer, process and service management, networking, packages, and power control, with optional multi-host switching. Operates without SSH or any Linux-side installation beyond one stdlib-only Python daemon, using a state-free binary protocol.

Enables an agent to verify its own work mid-task by deliberately breaking each changed line, re-running the test suite in isolated workers, and reporting which breaks no test catches, along with a plan tool that previews which changed lines would be checked.

Enables an AI agent to see and control the screen of an already-running macOS virtual machine over VNC, exposing four tools: screenshot, click, type_text, and key_press. This lets the agent drive a disposable VM's UI directly with its own model, without touching the real desktop or running a separate agent loop.

Enables MCP clients to query and diagnose machine-learning training runs through read-only tools for listing experiments, reading configurations and final metrics, comparing runs, sampling metric curves, and detecting training failures.

Enables read-only monitoring and diagnostics of Apache Flink clusters through a JobManager REST endpoint, including job, checkpoint, backpressure, TaskManager, and multi-cluster overviews with LLM-friendly summaries.

Provides OAuth-secured MCP tools for working with private, Git-backed Markdown knowledge bases: searching pages, reading content, ingesting new source documents, retrieving cited original files, and reviewing history or rolling back accepted pages. Every tool call is checked against each person's or agent's current wiki grant and returns source paths and Git revisions for traceability.

Exposes four read-only tools over extracted Hacker News "Who is hiring?" job posts — postings_search, posting_get, skills_top, and skills_trend — so an MCP client or agent can search postings, look up individual posts, rank top skills, or chart skill trends over time. Tools return only structured extracted fields and HN links (never raw post text) from a read-only SQLite database, including salary sorting and a hard row limit to keep queries cheap.

Enables AI agents to store decisions as a versioned history, linking each new decision to the one it replaces so that recall always resolves to the decision currently in force instead of obsolete ones. It runs fully locally on a single SQLite file with BM25 search, and can optionally detect replacements automatically via an LLM or heuristic linker.