32 lines
1.6 KiB
Markdown
32 lines
1.6 KiB
Markdown
---
|
||
type: project
|
||
status: active
|
||
domain: [ai, software, infra]
|
||
tags: [project, ai, agents]
|
||
repos: [clawd, clawdbot, clawd-dashboard, OpenClawDashboard, tiger-build]
|
||
---
|
||
# 🐾 Clawd Ecosystem — Self-Hosted AI Agent Platform
|
||
> A family of repos building an **autonomous AI agent** with a managed dashboard: memory, a skills/MCP registry, and an agent control plane.
|
||
|
||
## Components
|
||
| Repo | Role |
|
||
|---|---|
|
||
| **clawd** | Core agent (initial commit Feb 2026); "proved several architectural patterns" |
|
||
| **clawdbot** | Bot interface |
|
||
| **clawd-dashboard** | Active dashboard (24 commits, Feb–Jun 2026) |
|
||
| **OpenClawDashboard** | "Clawd Agent Dashboard" — memory mgmt (`MEMORY.md`), **Skills Registry**, MCP management (7 commits) |
|
||
| **tiger-build** | Agent UI build — activity feed, agent chips, agent-file routes, model config (large component set) |
|
||
|
||
## Purpose
|
||
Build and operate a personal AI agent with: editable **core memory**, a browsable **skills/capabilities registry**, **MCP** server management, and an activity/observability dashboard.
|
||
|
||
## Why it matters
|
||
This is Manohar's deepest **AI-agent infrastructure** work — directly informs [[Agent Architectures]], [[MCP]], and [[MOC - AI]]. The memory-file pattern here echoes how *this very vault's* [[Decision Log]] is structured.
|
||
|
||
## Open questions
|
||
- [ ] Is "Clawd" a play on Claude? Is this built on the Claude Agent SDK / Claude Code?
|
||
- [ ] Which of the 5 repos is canonical/current ([[tiger-build]] vs clawd-dashboard)?
|
||
- [ ] Intended as a product or personal infra?
|
||
|
||
## Related
|
||
[[Agent Architectures]] · [[Agent Frameworks]] · [[MCP]] · [[Cortex]] · [[MOC - AI]]
|