{"id":32023582,"url":"https://github.com/defai-digital/automatosx","last_synced_at":"2026-04-02T11:57:16.893Z","repository":{"id":318493840,"uuid":"1071156967","full_name":"defai-digital/AutomatosX","owner":"defai-digital","description":"AutomatosX is an orchestrates AI agents, workflows, and memory","archived":false,"fork":false,"pushed_at":"2026-03-21T08:08:50.000Z","size":24206,"stargazers_count":21,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"main","last_synced_at":"2026-03-21T22:26:56.646Z","etag":null,"topics":["agentic-framework","ai-automation","ai-runtime","aicoding","claude-code","codex-cli","developer-tools","gemini-cli","glm","grok","multi-agent","open-source","openai","opencode","qwen","workflow-orchestration"],"latest_commit_sha":null,"homepage":"https://automatosx.com","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/defai-digital.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-10-07T00:21:28.000Z","updated_at":"2026-03-21T08:08:40.000Z","dependencies_parsed_at":"2026-01-09T05:04:16.386Z","dependency_job_id":null,"html_url":"https://github.com/defai-digital/AutomatosX","commit_stats":null,"previous_names":["defai-sg/automatosx","defai-digital/automatosx"],"tags_count":338,"template":false,"template_full_name":null,"purl":"pkg:github/defai-digital/AutomatosX","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/defai-digital%2FAutomatosX","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/defai-digital%2FAutomatosX/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/defai-digital%2FAutomatosX/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/defai-digital%2FAutomatosX/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/defai-digital","download_url":"https://codeload.github.com/defai-digital/AutomatosX/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/defai-digital%2FAutomatosX/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31305966,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-02T09:48:21.550Z","status":"ssl_error","status_checked_at":"2026-04-02T09:48:19.196Z","response_time":89,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agentic-framework","ai-automation","ai-runtime","aicoding","claude-code","codex-cli","developer-tools","gemini-cli","glm","grok","multi-agent","open-source","openai","opencode","qwen","workflow-orchestration"],"created_at":"2025-10-16T09:52:54.261Z","updated_at":"2026-04-02T11:57:16.886Z","avatar_url":"https://github.com/defai-digital.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# AutomatosX OSS\n\n**AI orchestration through MCP - Supercharge your AI coding assistant**\n\n[![Version](https://img.shields.io/badge/version-14.0.0-green.svg)](https://github.com/defai-digital/AutomatosX/releases)\n[![Node.js](https://img.shields.io/badge/node-%3E%3D22.5.0-blue.svg)](https://nodejs.org)\n[![License](https://img.shields.io/badge/license-BSL--1.1-blue.svg)](LICENSE)\n\n---\n\n## What is AutomatosX?\n\nAutomatosX adds **80+ specialized tools** to your AI coding assistant through MCP (Model Context Protocol). Instead of just chatting with Claude, Gemini, or Codex, you get:\n\n- **28 specialized agents** (fullstack, security, devops, ml-engineer, meta-agents, etc.)\n- **Multi-model discussions** (ask Claude AND Gemini AND Grok the same question)\n- **Code review with focus areas** (security, performance, architecture)\n- **Persistent memory** across sessions\n- **Governance gates** to check AI-generated code\n- **Execution tracing** for full audit trails\n- **Workflow engine** with delegate steps, guard policies, and retry logic\n- **SQLite-backed stores** for memory, traces, and state\n\n---\n\n## Quick Start (3 Steps)\n\n```bash\n# 1. Install AutomatosX\nnpm install -g @defai.digital/cli\n\n# 2. Global setup (one-time)\nax setup\n\n# 3. Initialize your project (registers MCP with your AI CLI)\ncd your-project\nax init\n```\n\nNow use your AI CLI - it has access to all AutomatosX tools.\n\n---\n\n## Using AutomatosX with Your AI CLI\n\nAfter running `ax init`, your AI assistant gains access to 80+ new tools. Here's how to use them:\n\n### With Claude Code\n\n```bash\n# Start Claude Code in your project\nclaude\n\n# Now you can use AutomatosX tools in conversation:\n```\n\n**Example conversations with Claude:**\n\n```\nYou: Use ax_agent_recommend to find the best agent for \"implement OAuth2 login\"\n\nClaude: [Uses ax_agent_recommend tool]\nThe recommended agent is \"security\" with 0.85 confidence because...\n\nYou: Run that agent with ax_agent_run\n\nClaude: [Uses ax_agent_run tool with agentId: \"security\"]\nHere's the OAuth2 implementation plan...\n```\n\n```\nYou: Use ax_discuss to get multiple AI perspectives on \"REST vs GraphQL for mobile app\"\n\nClaude: [Uses ax_discuss tool]\nAfter synthesizing perspectives from Claude, Gemini, and Grok:\n- REST is recommended for... [consensus summary]\n```\n\n```\nYou: Run ax_review_analyze on src/api/ with focus on security\n\nClaude: [Uses ax_review_analyze tool]\nFound 3 security issues:\n1. SQL injection risk in users.ts:45...\n```\n\n---\n\n### With Gemini CLI\n\n```bash\ngemini\n```\n\n```\nYou: I need to build a caching layer. Use ax_agent_run with the architecture agent.\n\nGemini: [Uses ax_agent_run tool]\nHere's a recommended caching architecture...\n\nYou: Now use ax_discuss_quick to validate this approach with other models\n\nGemini: [Uses ax_discuss_quick tool]\nMulti-model consensus: The proposed approach is sound, with these refinements...\n```\n\n---\n\n### With Codex CLI\n\n```bash\ncodex\n```\n\n```\nYou: Use ax_agent_list to show me available agents\n\nCodex: [Uses ax_agent_list tool]\nAvailable agents (28 total):\n- fullstack: Senior Fullstack Engineer\n- backend: Backend Engineer\n- security: Security Engineer\n- architect: Strategic Technology Leader (meta-agent)\n...\n```\n\n---\n\n### With Grok (via ax-cli)\n\n```bash\nax-grok\n```\n\n```\nYou: Use ax_guard_check with policy \"bugfix\" on the files I changed\n\nGrok: [Uses ax_guard_check tool]\nGuard check passed: All changes within policy boundaries.\n```\n\n---\n\n### With Cursor IDE\n\nCursor IDE automatically detects the `.cursor/mcp.json` config created by `ax init`.\n\n1. Run `ax init` in your project\n2. Open the project in Cursor\n3. Use AutomatosX tools in Cursor's AI chat\n\n---\n\n## Available MCP Tools (80+ total)\n\n### Agent Tools\n| Tool | Description |\n|------|-------------|\n| `ax_agent_list` | List all specialized agents |\n| `ax_agent_run` | Execute an agent with input |\n| `ax_agent_get` | Get agent details |\n| `ax_agent_recommend` | Find best agent for a task |\n| `ax_agent_register` | Create custom agent |\n| `ax_agent_remove` | Remove an agent |\n| `ax_agent_capabilities` | List all capabilities |\n\n### Discussion Tools\n| Tool | Description |\n|------|-------------|\n| `ax_discuss` | Multi-model discussion with synthesis |\n| `ax_discuss_quick` | Quick 2-round consensus |\n| `ax_discuss_recursive` | Multi-level recursive discussions with depth control |\n\n### Review Tools\n| Tool | Description |\n|------|-------------|\n| `ax_review_analyze` | Code review with focus (security, performance, architecture, etc.) |\n| `ax_review_list` | List recent reviews |\n\n### Guard Tools\n| Tool | Description |\n|------|-------------|\n| `ax_guard_check` | Check changes against policy |\n| `ax_guard_list` | List available policies |\n| `ax_guard_apply` | Apply policy to session |\n\n### Memory Tools\n| Tool | Description |\n|------|-------------|\n| `ax_memory_store` | Store key-value data |\n| `ax_memory_retrieve` | Retrieve stored data |\n| `ax_memory_search` | Search memory |\n| `ax_memory_list` | List all keys |\n| `ax_memory_delete` | Delete a key |\n\n### Session Tools\n| Tool | Description |\n|------|-------------|\n| `ax_session_create` | Create collaboration session |\n| `ax_session_join` | Join existing session |\n| `ax_session_complete` | Mark session complete |\n| `ax_session_status` | Check session status |\n| `ax_session_list` | List sessions |\n| `ax_session_leave` | Leave a session |\n| `ax_session_fail` | Mark session failed |\n| `ax_session_close_stuck` | Close stuck sessions |\n\n### Workflow Tools\n| Tool | Description |\n|------|-------------|\n| `ax_workflow_run` | Execute a workflow |\n| `ax_workflow_list` | List workflows |\n| `ax_workflow_describe` | Get workflow details |\n\n### Trace Tools\n| Tool | Description |\n|------|-------------|\n| `ax_trace_list` | List execution traces |\n| `ax_trace_get` | Get trace details |\n| `ax_trace_analyze` | Analyze trace for issues |\n| `ax_trace_tree` | Get hierarchical trace tree |\n| `ax_trace_by_session` | Get traces for a session |\n| `ax_trace_close_stuck` | Close stuck traces |\n\n### Scaffold Tools\n| Tool | Description |\n|------|-------------|\n| `ax_scaffold_contract` | Generate Zod schema |\n| `ax_scaffold_domain` | Generate domain package |\n| `ax_scaffold_guard` | Generate guard policy |\n\n### Config Tools\n| Tool | Description |\n|------|-------------|\n| `ax_config_get` | Get config value |\n| `ax_config_set` | Set config value |\n| `ax_config_show` | Show full config |\n\n### File System Tools\n| Tool | Description |\n|------|-------------|\n| `ax_file_write` | Write file content |\n| `ax_file_exists` | Check if file exists |\n| `ax_directory_create` | Create directory |\n\n### Ability Tools\n| Tool | Description |\n|------|-------------|\n| `ax_ability_list` | List available abilities |\n| `ax_ability_inject` | Inject abilities into prompt |\n\n### Parallel Execution Tools\n| Tool | Description |\n|------|-------------|\n| `ax_parallel_run` | Execute multiple agents in parallel with DAG dependencies |\n| `ax_parallel_plan` | Preview execution plan without running |\n\n### Semantic Search Tools\n| Tool | Description |\n|------|-------------|\n| `ax_semantic_store` | Store content with vector embeddings |\n| `ax_semantic_search` | Find similar content by meaning |\n| `ax_semantic_get` | Retrieve specific item by key |\n| `ax_semantic_list` | List stored items |\n| `ax_semantic_delete` | Remove item from store |\n| `ax_semantic_stats` | Storage statistics |\n| `ax_semantic_clear` | Clear namespace |\n\n### Research Tools\n| Tool | Description |\n|------|-------------|\n| `ax_research_query` | Web search with AI synthesis |\n| `ax_research_fetch` | Fetch and extract from URL |\n| `ax_research_synthesize` | Combine sources into answer |\n\n### Design Tools\n| Tool | Description |\n|------|-------------|\n| `ax_design_api` | Generate OpenAPI/AsyncAPI specs |\n| `ax_design_component` | Create component interface designs |\n| `ax_design_schema` | Generate Zod/JSON schemas |\n| `ax_design_architecture` | Create architecture diagrams (Mermaid, PlantUML, C4) |\n| `ax_design_list` | List design artifacts |\n\n### Git Tools\n| Tool | Description |\n|------|-------------|\n| `ax_git_status` | Repository status |\n| `ax_git_diff` | Show file changes |\n| `ax_commit_prepare` | Stage files and generate commit message |\n| `ax_pr_create` | Create GitHub pull request with AI description |\n| `ax_pr_review` | Get PR details for review |\n\n### Feedback Tools\n| Tool | Description |\n|------|-------------|\n| `ax_feedback_submit` | Submit task feedback |\n| `ax_feedback_history` | View feedback history |\n| `ax_feedback_stats` | Agent feedback statistics |\n| `ax_feedback_overview` | System-wide feedback summary |\n| `ax_feedback_adjustments` | View score adjustments |\n\n### MCP Ecosystem Tools\n| Tool | Description |\n|------|-------------|\n| `ax_mcp_server_register` | Register external MCP server |\n| `ax_mcp_server_list` | List registered servers |\n| `ax_mcp_server_unregister` | Remove server registration |\n| `ax_mcp_tools_discover` | Discover tools from servers |\n| `ax_mcp_tool_invoke` | Call tool on external server |\n| `ax_mcp_tools_list` | List discovered tools |\n\n---\n\n## Example Workflows\n\n### 1. Get Multiple AI Perspectives\n\n```\nYou: Use ax_discuss to analyze \"Should we use microservices or monolith for a 5-person startup?\"\n\nAI: [Calls ax_discuss with topic and default providers: claude, gemini, grok]\n\nResult: After 2 rounds of discussion, the consensus is:\n- Start with a modular monolith (recommended by all 3 models)\n- Claude emphasizes: Operational simplicity crucial for small teams\n- Gemini adds: Consider service boundaries for future extraction\n- Grok suggests: Use feature flags for gradual migration later\n```\n\n### 2. Security-Focused Code Review\n\n```\nYou: Run ax_review_analyze on src/auth/ with focus \"security\"\n\nAI: [Calls ax_review_analyze with paths: [\"src/auth/\"], focus: \"security\"]\n\nResult: Found 2 issues:\n1. HIGH: Potential SQL injection in src/auth/login.ts:45\n   - User input concatenated directly in query\n   - Recommendation: Use parameterized queries\n\n2. MEDIUM: Missing rate limiting on src/auth/reset-password.ts\n   - Endpoint vulnerable to brute force\n   - Recommendation: Add rate limiting middleware\n```\n\n### 3. Find the Right Agent for Your Task\n\n```\nYou: Use ax_agent_recommend for \"set up Kubernetes deployment with auto-scaling\"\n\nAI: [Calls ax_agent_recommend with task]\n\nResult:\n- Recommended: devops (confidence: 0.92)\n- Reason: Task involves K8s, deployment automation, infrastructure\n- Alternatives:\n  - backend (0.45) - can help with app configuration\n  - architecture (0.38) - can review scaling strategy\n```\n\n### 4. Execute a Specialized Agent\n\n```\nYou: Run ax_agent_run with agentId \"devops\" and input {\"query\": \"Create Kubernetes deployment for Node.js app with HPA\"}\n\nAI: [Calls ax_agent_run]\n\nResult: Here's your Kubernetes configuration:\n[Agent generates deployment.yaml, service.yaml, hpa.yaml with explanations]\n```\n\n### 5. Check Changes Against Policy\n\n```\nYou: Use ax_guard_check with policy \"bugfix\" and changedPaths from my git diff\n\nAI: [Calls ax_guard_check with policy: \"bugfix\", changedPaths: [\"src/api/users.ts\", \"src/utils/validate.ts\"]]\n\nResult:\n- Status: PASS\n- Changed packages: 2 (within limit of 3)\n- No forbidden paths modified\n- Dependency boundaries respected\n```\n\n---\n\n## Available Agents (28)\n\n### Meta-Agents (Dynamic Capability Loading)\n| Agent | Expertise | Best For |\n|-------|-----------|----------|\n| `architect` | System design, tech strategy, product | High-level architecture and planning |\n| `auditor` | Security, testing, compliance | Quality and security verification |\n| `builder` | Fullstack, backend, frontend, ML | Implementation orchestration |\n| `executor` | Task routing, workflow execution | Complex multi-step tasks |\n| `operator` | DevOps, MLOps, deployment | Infrastructure and operations |\n| `researcher` | Analysis, data science, writing | Research and documentation |\n| `reviewer` | Code review, security, quality | Verification and validation |\n\n### Specialist Agents\n| Agent | Expertise | Best For |\n|-------|-----------|----------|\n| `fullstack` | React, Node.js, TypeScript | End-to-end features |\n| `backend` | APIs, databases, microservices | Server-side development |\n| `frontend` | React, Vue, CSS, accessibility | UI components |\n| `architecture` | System design, patterns | Architecture decisions |\n| `security` | OWASP, audits, secure coding | Security reviews |\n| `devops` | Docker, K8s, CI/CD | Infrastructure |\n| `mobile` | React Native, Flutter | Mobile apps |\n| `blockchain-developer` | Solidity, DeFi, NFTs | Web3 development |\n| `data-scientist` | Python, ML, statistics | Data analysis |\n| `ml-engineer` | PyTorch, MLOps | Production ML |\n| `mlops-engineer` | ML pipelines, deployment | ML infrastructure |\n| `bug-hunter` | Bug detection, edge cases | Proactive bug hunting |\n| `cto` | Strategy, leadership | Technical direction |\n| `ceo` | Business strategy | Business decisions |\n| `product` | Requirements, roadmaps | Product planning |\n| `quantum-engineer` | Quantum algorithms | Quantum computing |\n| `aerospace-scientist` | Aerospace systems | Aerospace engineering |\n| `creative-marketer` | Marketing, content | Marketing campaigns |\n| `quality` | Testing, QA | Test planning |\n| `writer` | Documentation, API docs | Technical writing |\n| `standard` | General assistance | Simple tasks |\n\n---\n\n## CLI Commands\n\n```bash\n# System\nax setup                    # Global setup (one-time)\nax init                     # Project initialization (per-project)\nax doctor                   # Check provider health\nax status                   # Runtime status\nax monitor                  # Launch web dashboard\n\n# Direct provider calls\nax call claude \"Explain this code\"\nax call gemini --file ./src/api.ts \"Review this\"\nax call --autonomous --goal \"refactor auth module\" --max-rounds 5\n\n# Agents\nax agent list\nax agent run security --input '{\"query\": \"audit auth\"}'\n\n# Review\nax review analyze src/ --focus security\nax review analyze src/ --since main\n\n# Discussion\nax discuss \"REST vs GraphQL\"\n\n# Guard\nax guard check --policy bugfix --changed-paths src/\n\n# Workflows\nax run \u003cworkflow-id\u003e\nax ship --scope \u003carea\u003e\nax architect --request \"\u003crequirement\u003e\"\nax audit --scope \u003cpath\u003e\nax qa --target \u003cservice\u003e --url \u003curl\u003e\nax release --release-version \u003cversion\u003e\n\n# Traces\nax trace analyze \u003ctrace-id\u003e\nax trace by-session \u003csession-id\u003e\nax trace tree \u003ctrace-id\u003e\n\n# Other\nax ability list\nax feedback submit\nax iterate \u003ccommand\u003e --max-rounds 3\nax resume \u003ctrace-id\u003e\nax history\nax scaffold contract\nax update\n```\n\n---\n\n## Workflow Engine (v14)\n\nv14 ships a full workflow execution engine with support for all step types:\n\n| Step Type | Description |\n|-----------|-------------|\n| `prompt` | LLM prompt execution |\n| `tool` | MCP tool invocation |\n| `conditional` | Branch on context values |\n| `loop` | Iterate over collections |\n| `parallel` | Concurrent step execution |\n| `discuss` | Multi-model discussion step |\n| `delegate` | Route to a registered agent (with depth + circular guards) |\n\nWorkflows are defined as YAML files and executed via `ax run` or `ax_workflow_run`.\n\n---\n\n## Provider Installation\n\nInstall at least one AI provider CLI:\n\n| Provider | Command | Installation |\n|----------|---------|--------------|\n| Claude | `claude` | [Claude Code](https://github.com/anthropics/claude-code) |\n| Gemini | `gemini` | [Gemini CLI](https://github.com/google-gemini/gemini-cli) |\n| Codex | `codex` | [Codex CLI](https://github.com/openai/codex) |\n| Grok | `ax-grok` | [ax-cli](https://github.com/defai-digital/ax-cli) (requires XAI_API_KEY) |\n| OpenCode | `opencode` | [OpenCode](https://github.com/opencode-ai/opencode) |\n\n---\n\n## IDE Integration\n\n| IDE | MCP Config Location |\n|-----|---------------------|\n| [Cursor](https://cursor.com/) | `.cursor/mcp.json` (per-project) |\n| [Antigravity](https://antigravity.google) | `~/.gemini/antigravity/mcp_config.json` (global) |\n\n`ax init` automatically configures MCP for all detected providers and IDEs.\n\n---\n\n## Setup vs Init\n\n- **`ax setup`** — Global, one-time setup. Detects installed provider CLIs, creates global config at `~/.automatosx/`.\n- **`ax init`** — Per-project initialization. Creates `.automatosx/` in current directory, registers MCP server with all detected providers and IDEs.\n\n---\n\n## What's New in v14\n\n- **Monorepo architecture** — Split into focused packages: `contracts`, `workflow-engine`, `shared-runtime`, `state-store`, `trace-store`, `mcp-server`, `monitoring`, `cli`\n- **Workflow engine** — Full step type support including `delegate` with depth and circular delegation guards\n- **SQLite backends** — Native `node:sqlite` storage for state-store and trace-store (WAL mode, FTS5 full-text search)\n- **Unified runtime** — CLI and MCP share a single `shared-runtime`; no duplicated orchestration logic\n- **New CLI commands** — `ability`, `feedback`, `history`, `iterate`, `monitor`, `scaffold`, `update`\n- **Stronger guards** — Built-in `safe-filesystem` policy; secret leakage detection; prototype pollution protection\n- **Legacy MCP aliases** — `ax_*` tool names from v13.5 continue to work\n\n[Full Changelog](CHANGELOG.md)\n\n---\n\n## License\n\n**Business Source License 1.1** - see [LICENSE](LICENSE)\n\n- **Effective starting v13.3.0**: Source-available under BSL 1.1\n- **Free for non-production use**: Development, testing, personal projects\n- **Free for small production**: Organizations with \u003c $2M annual revenue\n- **Commercial license required**: Production use by organizations with \u003e= $2M annual revenue\n- **Converts to Apache 2.0**: Four years after each version's release date\n\nFor commercial licensing inquiries, contact: licensing@defai.digital\n\nSee [COMMERCIAL-LICENSE.md](COMMERCIAL-LICENSE.md) for full details\n\n---\n\n[GitHub](https://github.com/defai-digital/AutomatosX) | [Issues](https://github.com/defai-digital/AutomatosX/issues)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdefai-digital%2Fautomatosx","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdefai-digital%2Fautomatosx","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdefai-digital%2Fautomatosx/lists"}