Shape agent capabilities into your workflow.

Claude Engineer provides CLI and web interfaces where Claude can identify a missing capability, generate the corresponding tool code, load it dynamically, and keep using it during development work. It also manages persistent conversations, context and token usage, tool chaining, and debugging output, with optional Python execution through an E2B sandbox. Its core workflow depends on Anthropic Claude 3.5, and some tools require additional service credentials.

GitHub Copilot CLI brings its coding agent into a local command-line workflow for building, editing, debugging, and refactoring code. It can also work with repositories, issues, and pull requests through an authenticated GitHub account and extend its tools through MCP servers. Actions are previewed for approval, and use requires an active Copilot subscription with prompts counting against premium-request quotas.

Hive turns an objective into a graph-based runtime that coordinates specialized agents in parallel while preserving role memory, checkpoints, and execution state. It adds human-in-the-loop controls, observability, cost limits, browser and compute tools, and LiteLLM-compatible hosted or local models; teams building only simple one-off chains may not need this heavier harness.

Paseo runs a local daemon that manages Claude Code, Codex, Copilot, OpenCode, and Pi agents, then exposes them through desktop, mobile, web, and command-line clients. You can run agents in parallel on your own machines and pair a phone by QR code or use the headless server path. At least one supported agent CLI must already be installed and authenticated.

MCP Inspector connects to servers over stdio, SSE, or Streamable HTTP and lets developers inspect tools, resources, prompts, and responses in a browser UI. Its CLI supports scripted checks and continuous integration, and server configurations can be exported for other clients. Because the local proxy can spawn processes and connect to specified servers, it should not be exposed to untrusted networks.

Claurst is an open-source Rust terminal coding agent with a TUI, multiple model providers, plugins, chat forking, memory consolidation, headless use, and ACP editor integration. It can work on code through an agent tool loop without telemetry. The core agent and routing are in beta, while sustained goals and its highest-effort multi-agent workflow remain experimental.

visual-explainer is an agent skill that renders architecture explanations, diff reviews, plan audits, data tables, and project recaps as self-contained HTML pages or slide decks. It replaces fragile ASCII diagrams and terminal tables with readable typography, themes, and browser-based interaction. The repository provides installation or rules-based integration paths for Claude Code, Pi, Codex, OpenCode, and Cursor, depending on what each harness supports.

Kimi CLI combines a terminal agent with a Ctrl-X shell mode, a VS Code extension, ACP support for compatible editors, Zsh integration, and MCP server management. The project is being replaced by the next-generation Kimi Code CLI, which migrates configuration and sessions during installation. Existing installations remain available, but new users seeking ongoing updates should choose Kimi Code CLI.

PraisonAI supports both single-agent and multi-agent workflows for research, code generation, content work, data tasks, and cross-channel automation. Developers can start with the Python SDK or CLI, then add a visual flow builder, chat UI, or messaging dashboard as needed. Memory, knowledge retrieval, planning, MCP, agent handoffs, and broad model support are built in, while some web-search and messaging features require additional API keys or platform tokens.

Desktop Commander MCP equips clients such as Claude with local terminal execution, filesystem search and editing, process sessions, long-running commands, and support for formats including Excel, PDF, and DOCX. It can be installed through npx or run in Docker. The server is not itself a security sandbox: allowed-directory settings restrict file tools but do not fully contain terminal commands. Docker isolation and narrowly mounted folders are therefore important when working with untrusted tasks.

CAD Skills is an installable library of agent workflows for mechanical design, fabrication, and robotics. Its skills can create or edit CAD models from text or image requests, preview local CAD and G-code files, find off-the-shelf STEP parts, produce DXF drawings, prepare slicing work, and handle robot-description or simulation artifacts. STEP is the main modeling output, with supported exports such as STL, 3MF, and GLB. The library installs through the Skills CLI or provider-specific plugin paths for agents such as Codex and Claude Code.

Claude Squad runs coding agents such as Claude Code, Codex, Gemini, and Aider in separate Git worktrees, so several tasks can progress without colliding. Its tmux-based terminal interface keeps sessions alive, shows diffs for review, and lets you check out or push approved changes. tmux and GitHub CLI are required.

R2R ingests documents, images, audio, and other content into a self-hosted knowledge base, then exposes hybrid search, cited RAG, knowledge graphs, and a multi-step research agent through a REST API. Python and JavaScript clients handle document management and retrieval. It can run in a lightweight mode or a fuller Docker deployment, and normally requires an API key for the selected language model.

nullclaw packages model providers, messaging channels, memory backends, tools, and service management into a self-contained Zig binary for ARM, x86, and RISC-V systems. It can run as an interactive agent, one-shot command, gateway, or background service, with implementations selected through configuration. The project reports a roughly 678 KB binary and about 1 MB peak memory use.

Claude Agent SDK for Python exposes an async query function and ClaudeSDKClient for bidirectional, interactive conversations with Claude Code. Custom tools and hooks can run as in-process MCP servers, while external MCP servers can be used alongside them. The package bundles the Claude Code CLI by default, requires Python 3.10 or newer, and allows a custom CLI path.

camofox-browser wraps Camoufox in a REST service built for agent browsing, using compact accessibility snapshots and stable element references for navigation, clicking, typing, and extraction. It isolates cookies and storage by user and supports proxies, imported login state, traces, and several deployment targets. The first install downloads roughly 300 MB, and anonymized crash telemetry is enabled by default but can be disabled.

Refly turns business procedures and team SOPs into reusable, versioned Agent Skills through a visual workflow builder. Skills can be installed, published, and shared through a registry, then exposed to other systems as APIs, webhooks, MCP services, or portable SDKs. It is aimed at teams that want durable agent capabilities rather than one-off prompts or copied automation templates.

Use compatible coding agents to create or edit Google Stitch screens, import existing frontend code into a design, manage design systems, and generate implementation assets. The collection includes workflows for React, React Native, and other build targets, but the design operations depend on the Stitch MCP server and selectively installed skills may require companion skills.