Agent Ha Bridge — AI skill for Claude Code
Answer your AI coding agent from Home Assistant or your terminal — GitHub Copilot CLI, Claude Code, Codex CLI, or any MCP client.
How to install Agent Ha Bridge
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open danswett/agent-ha-bridge and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agent Ha Bridge does
Answer your AI coding agent from Home Assistant or your terminal — GitHub Copilot CLI, Claude Code, Codex CLI, or any MCP client. Per-session dashboard cards, live activity, chain-of-thought, and answers typed straight back into the terminal.
Alternatives in AI
- Tokentap — Intercept LLM API traffic and visualize token usage in a real-time terminal dashboard 763 ★
- Prismor — Self-hosted runtime control plane for AI agents 334 ★
- Openclaw MCP — 🦞 MCP server for OpenClaw - secure bridge between Claude.ai and your self-hosted OpenClaw assistant with OAut 185 ★
README
AI coding agent ⇄ Home Assistant bridge
[](https://github.com/danswett/agent-ha-bridge/actions/workflows/ci.yml) [](https://github.com/danswett/agent-ha-bridge/actions/workflows/codeql.yml)
Answer your AI coding agent from Home Assistant — or from your terminal — whichever you happen to be looking at. Works with **GitHub Copilot CLI**, **Claude Code**, **OpenAI Codex CLI**, and any **MCP client**.
Every live session gets its own card on a Home Assistant dashboard showing what it's doing, what it just said, and what it's waiting on. When the agent asks a question, the card grows the matching controls. Whatever you pick is typed into the real terminal prompt, so the terminal never stops working and nothing is ever answered twice.
**Windows only** for the terminal integration. Answers are delivered into the running CLI with `AttachConsole` + `WriteConsoleInput`, which is Win32-specific. Everything else is portable — and the [MCP server](mcp/) needs no console injection at all, so it runs on any OS.
What you get
| Dual input | Answer in the terminal or Home Assistant. First one wins; the other clears. |
| Live activity | Each session streams its status, current tool, and last response to its card. |
| Detailed activity | A dashboard toggle streams the model's reasoning and every tool call into an expander. |
| Real forms | Multi-field questions become one dropdown per field plus a Send button. |
| Continuation | Reply to a finished turn from your phone; it's typed into the session. |
| Start a conversation | A session appears as soon as it opens, so you can send it its first prompt from the dashboard. |
| Launch a session | Pick a workspace, type an opening prompt, press a button — a new CLI session opens on your desktop. |
Related Skills
LLMs In Practice For Students
Practical LLM course for engineering students — prompt engineering, few-shot learning, chain-of-thought, and b
Sidekick Agent Hub
See what your AI coding agent is doing. Multi-provider assistant & session monitor for VS Code and the termina
Che Dashboard Agent
A containerized AI coding assistant for the Eclipse Che Dashboard's Devfile Creator. Runs Claude Code inside a
Falaq Cockpit OSS
Mission control for Claude Code & Codex — a local dashboard to monitor, multitask, and drive all your AI codin
Dashboard Jira
Local board that orchestrates Claude Code sessions from Jira tickets: approve cards from desktop or phone and
HestiaClaw
The AI that remembers, learns, and actually does things in your Home Assistant environment.
Related Agents
Git Analytics
Use to audit recent design-system change activity from LOCAL git history: breaking / additive / cosmetic chang
Nvidia Cuda Engineer
You are the system GPU Engineer — a senior NVIDIA platform specialist embedded in the system Distributed AI Ho
Wsl Worker
Handle git and file operations for the WSL repository /home/ljq/code/PINN/SolvingTeukolsky through the fixed b