Polyglot-agent dashboard in statically typed Common Lisp (TUI, Web, McCLIM)
- Common Lisp 97.2%
- JavaScript 1.3%
- Shell 0.9%
- TypeScript 0.3%
- Makefile 0.2%
| .beads | ||
| .github/workflows | ||
| artifacts | ||
| attic | ||
| ci | ||
| docs | ||
| e2e | ||
| e2e-tui | ||
| packaging | ||
| scripts | ||
| src | ||
| test-harness | ||
| tests | ||
| .gitignore | ||
| agent-orrery.asd | ||
| AGENTS.md | ||
| CHANGELOG.md | ||
| Makefile | ||
| README.md | ||
| run-tui-tests.lisp | ||
Agent Orrery
Agent Orrery is a public Kyvero Vexus project to provide a multi-interface operations dashboard for OpenClaw-compatible agent systems.
Like an orrery models celestial bodies from multiple viewpoints, Agent Orrery models agentic runtime state across:
- Terminal UI (TUI)
- Web UI
- McCLIM UI
The platform is OpenClaw-compatible first, but built around a generic adapter protocol so it can support other agentic systems with minimal glue code.
Core Design Commitments
- Statically typed Common Lisp
- Coalton for pure core logic
- SBCL strict declarations for integration/runtime layers
- Functional programming style first (immutable data flows, pure transforms, side effects at boundaries)
- Modular architecture (core domain + adapter boundary + UI frontends)
- Extensible provider model (OpenClaw adapter is just the first implementation)
- E2E test-gated delivery (every epic must land with real interaction tests)
Inspiration Sources
- https://github.com/tugcantopaloglu/openclaw-dashboard
- https://github.com/mudrii/openclaw-dashboard
- https://github.com/abhi1693/openclaw-mission-control
Planning Docs
- Full plan:
docs/PLAN.md - E2E matrix:
docs/E2E-MATRIX.md - Architecture:
docs/ARCHITECTURE.md
Task Tracking
This repository uses beads (bd) for all issue tracking.
bd ready --json