Zerwiz

Agent OS fills the gap when AI coding agents lack your project context. It discovers and documents coding standards from your codebase, then injects them into agent context. Works with Claude Code, Cursor, Windsurf, and other AI tools. Core capabilities: Discover Standards, Deploy Standards, Shape Spec, Index Standards. Free & open source.

• ▸ Discover patterns from your codebase
• ▸ Inject standards into agent context
• ▸ Shape better specs for spec-driven development
• ▸ Works with Cursor, Claude Code, Windsurf

BMad agents and workflows act as expert collaborators who guide you through a structured process. Scale-domain-adaptive: adjusts planning depth based on project complexity. Structured workflows grounded in agile best practices. Specialized agents: PM, Architect, Developer, UX, Scrum Master, and more. Party Mode: multiple agent personas in one session. 100% free and open source. Install: npx bmad-method install

• ▸ 21+ specialized agents, 50+ workflows
• ▸ Scale-adaptive intelligence
• ▸ Quick Flow: /quick-spec, /dev-story, /code-review
• ▸ Full planning: product brief, PRD, architecture, epics
• ▸ Works with Cursor, Claude Code, Windsurf

Comprehensive framework for project and company operations. Templates, workflows, rules, and standards based on best practices. Helps structure projects so AI agents and teams stay aligned. Generate complete project structures, ensure compliance, integrate NorthStar tools. By NorthStar team via Nordic Global Trade. Want to help develop NSR? Request access on GitHub or contact us.

• ▸ Project and company operations templates
• ▸ Workflows, rules, and standards
• ▸ Generate complete project structures
• ▸ Compliance and best practices
• ▸ NorthStar team: Josef Lindbom, Craig Martin

Agent Zero is not a predefined agentic framework. It is designed to be dynamic, organically growing, and learning as you use it. Uses the computer as a tool: write code, run terminal, search, communicate. Skills System with portable SKILL.md standard (compatible with Claude Code, Cursor, Codex). Multi-agent cooperation with subordinate agents. Docker-ready, Web UI, LiteLLM. General-purpose assistant.

• ▸ Computer as tool—code execution, terminal, search
• ▸ Skills (SKILL.md standard, portable)
• ▸ Multi-agent cooperation
• ▸ Memory, projects, MCP support
• ▸ Docker: docker run -p 50001:80 agent0ai/agent-zero

The Pope Bot is an autonomous AI agent you can configure to do just about anything, 24/7. Secure: the AI cannot access your secrets; they are filtered at the process level. The repository IS the agent—every action is a git commit; audit and revert. Runs on GitHub Actions (free compute for public repos). Event handler creates jobs; Docker agent does the work and opens PRs. Telegram integration. npm run setup to get started.

• ▸ Secure by default—secrets never reach the LLM
• ▸ Repository IS the agent—every action a git commit
• ▸ Free compute via GitHub Actions
• ▸ Self-evolving through pull requests
• ▸ Telegram bot, webhook triggers

Pi is a minimal terminal coding agent—"there are many coding agents, but this one is mine." Powerful defaults, no baked-in sub-agents or plan mode; extend with extensions, skills, and packages. Four modes: interactive TUI, print/JSON, RPC, SDK. 15+ providers (Anthropic, OpenAI, Google, Ollama, OpenRouter, etc.). Tree-structured sessions, AGENTS.md/SYSTEM.md context, compaction, prompt templates. Extensions can add sub-agents, plan mode, MCP, sandboxing. Install: npm install -g @mariozechner/pi-coding-agent. MIT. By Mario Zechner & contributors.

• ▸ 15+ providers, hundreds of models; switch with /model or Ctrl+L
• ▸ Tree-structured sessions; /tree, /export, /share to gist
• ▸ AGENTS.md, SYSTEM.md, compaction, skills, prompt templates
• ▸ Extensions: TypeScript, tools, events; pi packages via npm or git
• ▸ Interactive, print/JSON, RPC, SDK modes; clawdbot integration example