orchestrator

# Orchestrator A lightweight, Python-based CLI tool and library for protocol-agnostic agent orchestration—combining MCP (Model Context Protocol) and A2A (Agent-to-Agent) support with GPT-OSS reasoning. ## Features - Dual-protocol orchestration: MCP for tools, A2A for agent collaboration - Intelligent reasoning engine powered by GPT-OSS (Hugging Face Inference) - Flexible CLI with mode flags (`--faves`, `--useful`, `--essential`), keyword filters (`--words`), and alias (`ts`) - AsyncIO-based drivers for STDIO/SSE (MCP) and HTTP/JSON-RPC (A2A) - Workflow engine with dependency management, retries, and state persistence - Enterprise-grade security: OAuth2/OpenID Connect, token management, TLS - Configurable via Pydantic models (YAML/JSON) and environment variables - Comprehensive testing, CI/CD, Docker packaging, and examples ## Installation Requirements: Python ≥ 3.11, `uv` environment manager ```bash # Clone repository git clone https://github.com/yourorg/orchestrator.git cd orchestrator # Initialize `uv` environment and install dependencies uv init uv install # Activate environment uv shell ``` ## Configuration Copy `examples/config.yaml` to `~/.ts/config.yaml` and update: ```yaml hf_api_key: YOUR_HUGGINGFACE_TOKEN gpt_oss_model: openai/gpt-oss-20B mcp_servers: - name: github command: ["npx", "@modelcontextprotocol/server-github"] env: GITHUB_TOKEN: "YOUR_GITHUB_TOKEN" - name: shortcut command: ["npx", "@shortcut/mcp"] env: SHORTCUT_API_TOKEN: "YOUR_SHORTCUT_TOKEN" a2a_endpoints: - name: k8s-agent url: https://agent.example.com/jsonrpc auth: token: "YOUR_AGENT_TOKEN" ``` Environment variables: - `HF_API_KEY` — Hugging Face API token - `A2A_AUTH_TOKEN` — default token for A2A endpoints ## Usage Run the CLI via `ts`: ```bash # Quick query with mode flags ts q "Deploy microservice to AWS" -ef # Orchestrate with protocol and reasoning level ts orchestrate "Setup CI/CD pipeline" --protocols=mcp,a2a --reasoning=high # Discover A2A agents by keyword ts discover a2a --words=devops,security # Create and execute a workflow template ts workflow create "enterprise-deploy" -u ts workflow execute "enterprise-deploy" ``` For full command reference: ```bash ts --help ts q --help ts orchestrate --help ``` ## Examples See `examples/basic_usage.sh` and `examples/advanced_workflow.yaml` for real-world scenarios. ## Testing Run tests with `pytest`: ```bash uv run pytest ``` ## Contributing 1. Fork the repository 2. Create a feature branch `git checkout -b feature/your-feature` 3. Install dependencies and run tests 4. Commit and push 5. Open a Pull Request Please follow code style (`black`, `flake8`) and include tests for new features. ## License MIT © Chrono's Cyber Chronicles