OpenMesha
# OpenMesha (OM)
[](https://github.com/ANAMIZED/OpenMesha/actions/workflows/ci.yml)
[](LICENSE)
[](https://www.python.org/downloads/)
[](src/openmesha/mcp/)
[](src/openmesha/sdk/)
[](src/openmesha/cli.py)
[](src/openmesha/api/)
[](src/openmesha/payments/)
**“Autonomous Local Economic System”**
## 🚀 Live Demo
[](https://anamized.github.io/OpenMesha/OpenMesha-Briefing-SingleFile.html)
**Interactive 14-chapter systems briefing** (fully self-contained single file) — kernel, agents, economy, security, quantum, x402, HOTL, and more.
→ [Open the live interactive demo](https://anamized.github.io/OpenMesha/OpenMesha-Briefing-SingleFile.html)
---
OpenMesha is a production-oriented agentic operating system: Super kernel (Thompson router, bitemporal memory, QAOA/VQE, AgentFence, x402, HOTL, PQC) delivered as a self-contained web control plane, plus a Python control plane with OS-like primitives — agents as processes, budgets, multi-agent workflows, MCP, SDK, and CLI.
A senior engineer who has never seen this repository can, using **only** the source and this `README.md`:
1. Deploy the entire system (single command)
2. Exercise every major feature
3. Verify end-to-end correctness via automated checks
**[Support Agentic OS Kernels ($99)](https://buy.stripe.com/bJecN63wObPv6Bf7Zm43S02)** · **[Agentic OS Cycle ($0.75)](https://buy.stripe.com/3cI14o8R8dXD3p3frO43S04)** · **[Public Goods Support](https://donate.stripe.com/00w5kE3wOg5L8Jn2F243S00)**
### Non-custodial USDC (preferred for agents)
| Network | Address | Explorer |
|---------|---------|----------|
| **Base** | `0xD3d0E9eDAe3Ac7bb199a8EAA761BdA423b878438` | [basescan](https://basescan.org/address/0xD3d0E9eDAe3Ac7bb199a8EAA761BdA423b878438) |
| **Ethereum** | `0xD3d0E9eDAe3Ac7bb199a8EAA761BdA423b878438` | [etherscan](https://etherscan.io/address/0xD3d0E9eDAe3Ac7bb199a8EAA761BdA423b878438) |
| **Solana** | `ETQwWf19axArsY493UfC6bxe2BmEzmzvCb58PPnC38A` | [solscan](https://solscan.io/account/ETQwWf19axArsY493UfC6bxe2BmEzmzvCb58PPnC38A) |
*Related:* [rui](https://github.com/ANAMIZED/rui) · [server-os](https://github.com/ANAMIZED/server-os) · [LRSI](https://github.com/ANAMIZED/LRSI) · [OpenGOS](https://github.com/ANAMIZED/OpenGOS)
## Surfaces
| Surface | Entry |
|---------|-------|
| **Live Demo (GitHub Pages)** | [OpenMesha-Briefing-SingleFile.html](https://anamized.github.io/OpenMesha/OpenMesha-Briefing-SingleFile.html) |
| **Web control plane** | [`web/openmesha.html`](web/openmesha.html) |
| REST API | `openmesha-api` / `python -c "from openmesha.main import run; run()"` |
| CLI | `openmesha status` / `openmesha agents ...` / `openmesha workflow` |
| MCP Server | `openmesha-mcp` |
| SDK | `from openmesha.sdk import OpenMeshaClient` |
| Multi-agent workflows | `POST /v1/workflows` · MCP · CLI · `skills/multi-agent-workflow/` |
| Skills | `skills/*/SKILL.md` |
| CI | `.github/workflows/ci.yml` |
| AGENTS.md | Coding-agent contract at repo root |
| **WIRE production path** | [`docs/WIRE.md`](docs/WIRE.md) — full cloud/self-hosted path: architecture, om-ctl API, stack, Foundry economics, settlement, P0–P4 checklists |
## Quick Start
```bash
docker compose up --build
bash scripts/verify.sh
```
- API: http://localhost:8080
- OpenAPI: http://localhost:8080/docs
## Design principles
1. Least privilege by construction
2. Cost is a first-class resource
3. Fail closed (AgentFence + HOTL)
4. Honest offline simulation
5. Deployable with zero tribal knowledge
## License
Apache-2.0
TDQS
Scored across 9 tools
Each tool targets a distinct action or resource: agent lifecycle, workflow creation, reporting, capability listing, and payment quoting are clearly separated. The get_cost_ledger, get_audit_log, and get_metrics tools explicitly disambiguate their boundaries in their descriptions.
Most tools follow a clear verb_noun pattern such as list_agents, create_agent, run_task, and get_metrics. The outlier is x402_payment_quote, which is a noun-phrase name with no leading verb, creating a minor inconsistency.
Nine tools is a well-scoped count for an agent orchestration and governance server. Each tool addresses a meaningful part of the workflow without feeling redundant or padded.
The core lifecycle of creating agents, listing agents, running tasks, and creating workflows is covered, along with cost, audit, and metrics reporting. Missing update/delete/stop agent operations are notable but not critically blocking for the apparent in-process simulation scope.