@mseep/obsidian-agent-client 0.10.6

package/docs/.vitepress/theme/custom.css

@@ -0,0 +1,111 @@
+/**
+ * Obsidian-inspired theme for VitePress
+ * Based on Obsidian's default dark/light themes
+ */
+
+:root {
+  /* Obsidian Purple accent */
+  --vp-c-brand-1: #a88bfa;
+  --vp-c-brand-2: #8b6ce7;
+  --vp-c-brand-3: #7c5fd6;
+  --vp-c-brand-soft: rgba(168, 139, 250, 0.14);
+
+  /* Light mode colors (Obsidian light theme inspired) */
+  --vp-c-bg: #ffffff;
+  --vp-c-bg-alt: #f6f6f6;
+  --vp-c-bg-elv: #ffffff;
+  --vp-c-bg-soft: #f6f6f6;
+
+  --vp-c-text-1: #2e3338;
+  --vp-c-text-2: #5c6166;
+  --vp-c-text-3: #86909c;
+
+  --vp-c-divider: #e5e5e5;
+  --vp-c-border: #e0e0e0;
+  --vp-c-gutter: #e5e5e5;
+
+  /* Code blocks */
+  --vp-code-block-bg: #f6f6f6;
+  --vp-code-line-highlight-color: rgba(168, 139, 250, 0.1);
+}
+
+.dark {
+  /* Obsidian Dark mode colors */
+  --vp-c-brand-1: #a88bfa;
+  --vp-c-brand-2: #9575f3;
+  --vp-c-brand-3: #8466e5;
+  --vp-c-brand-soft: rgba(168, 139, 250, 0.16);
+
+  --vp-c-bg: #1e1e1e;
+  --vp-c-bg-alt: #181818;
+  --vp-c-bg-elv: #252525;
+  --vp-c-bg-soft: #252525;
+
+  --vp-c-text-1: #dcddde;
+  --vp-c-text-2: #a3a3a3;
+  --vp-c-text-3: #787878;
+
+  --vp-c-divider: #2d2d2d;
+  --vp-c-border: #3d3d3d;
+  --vp-c-gutter: #2d2d2d;
+
+  /* Code blocks */
+  --vp-code-block-bg: #181818;
+  --vp-code-line-highlight-color: rgba(168, 139, 250, 0.15);
+}
+
+/* Hero section styling */
+.VPHero .name {
+  background: linear-gradient(135deg, #a88bfa 0%, #7c5fd6 100%);
+  -webkit-background-clip: text;
+  background-clip: text;
+  -webkit-text-fill-color: transparent;
+}
+
+/* Button styling */
+.VPButton.brand {
+  background-color: var(--vp-c-brand-1);
+  border-color: var(--vp-c-brand-1);
+}
+
+.VPButton.brand:hover {
+  background-color: var(--vp-c-brand-2);
+  border-color: var(--vp-c-brand-2);
+}
+
+/* Sidebar active item */
+.VPSidebar .is-active > .link > .text {
+  color: var(--vp-c-brand-1);
+}
+
+/* Feature cards */
+.VPFeature {
+  border-radius: 8px;
+  transition: transform 0.2s ease, box-shadow 0.2s ease;
+}
+
+.VPFeature:hover {
+  transform: translateY(-2px);
+  box-shadow: 0 4px 12px rgba(0, 0, 0, 0.1);
+}
+
+.dark .VPFeature:hover {
+  box-shadow: 0 4px 12px rgba(0, 0, 0, 0.3);
+}
+
+/* Inline code */
+:not(pre) > code {
+  background-color: var(--vp-c-bg-soft);
+  border-radius: 4px;
+  padding: 2px 6px;
+  color: var(--vp-c-brand-1);
+}
+
+/* Custom containers (tips, warnings) */
+.custom-block.tip {
+  border-color: var(--vp-c-brand-1);
+}
+
+.custom-block.warning {
+  border-color: #e5a84b;
+}

package/docs/.vitepress/theme/index.ts

@@ -0,0 +1,4 @@
+import DefaultTheme from "vitepress/theme";
+import "./custom.css";
+
+export default DefaultTheme;

package/docs/agent-setup/claude-code.md

@@ -0,0 +1,84 @@
+# Claude Code Setup
+
+Claude Code is Anthropic's AI coding assistant. You can use it with either an **API key** or by **logging in with your Anthropic account**.
+
+## Install and Configure
+
+Open a terminal (Terminal on macOS/Linux, PowerShell on Windows) and run the following commands.
+
+1. Install claude-agent-acp:
+
+```bash
+npm install -g @agentclientprotocol/claude-agent-acp
+```
+
+2. Find the installation path:
+
+::: code-group
+
+```bash [macOS/Linux]
+which claude-agent-acp
+# Example output: /usr/local/bin/claude-agent-acp
+```
+
+```cmd [Windows]
+where.exe claude-agent-acp
+# Example output: C:\Users\Username\AppData\Roaming\npm\claude-agent-acp.cmd
+```
+
+:::
+
+3. Open **Settings → Agent Client**. The default command (`claude-agent-acp`) works in many cases. If the agent is not found automatically, set the **Claude Code path** to the path found above, or click **Auto-detect**.
+
+## Authentication
+
+Choose one of the following methods:
+
+### Option A: API Key
+
+1. Get your API key from [Anthropic Console](https://console.anthropic.com/)
+2. Enter the API key in **Settings → Agent Client → Claude Code → API key**
+
+### Option B: Account Login
+
+If you have a Claude subscription and prefer not to use an API key, you can log in with your Anthropic account.
+
+::: warning Important
+This requires installing **Claude Code CLI** separately. The CLI creates the login session that the plugin uses.
+:::
+
+1. Install Claude Code CLI by running the following command in your terminal:
+
+::: code-group
+
+```bash [macOS/Linux]
+curl -fsSL https://claude.ai/install.sh | bash
+```
+
+```powershell [Windows]
+irm https://claude.ai/install.ps1 | iex
+```
+
+:::
+
+2. Login via CLI by running:
+
+```bash
+claude
+```
+
+Follow the prompts to authenticate with your Anthropic account.
+
+3. In **Settings → Agent Client**, leave the **API key field empty**.
+
+::: tip
+The Claude Desktop app uses a different authentication system. Having Claude Desktop running does **not** authenticate the plugin — you must log in via the CLI.
+:::
+
+## Verify Setup
+
+1. Click the robot icon in the ribbon or use the command palette: **"Open chat view"**
+2. You should see the chat panel open and connect to Claude Code
+3. Try sending a message to verify the connection
+
+Having issues? See [Troubleshooting](/help/troubleshooting).

package/docs/agent-setup/codex.md

@@ -0,0 +1,76 @@
+# Codex Setup
+
+Codex is OpenAI's AI coding assistant. You can use it with either an **API key** or by **logging in with your OpenAI account**.
+
+## Install and Configure
+
+Open a terminal (Terminal on macOS/Linux, PowerShell on Windows) and run the following commands.
+
+1. Install codex-acp:
+
+```bash
+npm install -g @zed-industries/codex-acp
+```
+
+2. Find the installation path:
+
+::: code-group
+
+```bash [macOS/Linux]
+which codex-acp
+# Example output: /usr/local/bin/codex-acp
+```
+
+```cmd [Windows]
+where.exe codex-acp
+# Example output: C:\Users\Username\AppData\Roaming\npm\codex-acp.cmd
+```
+
+:::
+
+3. Open **Settings → Agent Client**. The default command (`codex-acp`) works in many cases. If the agent is not found automatically, set the **Codex path** to the path found above, or click **Auto-detect**.
+
+## Authentication
+
+Choose one of the following methods:
+
+### Option A: API Key
+
+1. Get your API key from [OpenAI Platform](https://platform.openai.com/api-keys)
+2. Enter the API key in **Settings → Agent Client → Codex → API key**
+
+### Option B: Account Login
+
+If you have a ChatGPT subscription and prefer not to use an API key, you can log in with your OpenAI account.
+
+::: warning Important
+This requires installing **Codex CLI** separately. The CLI creates the login session that the plugin uses.
+:::
+
+1. Install Codex CLI by running the following command in your terminal:
+
+```bash
+npm install -g @openai/codex
+```
+
+2. Login via CLI by running:
+
+```bash
+codex
+```
+
+Follow the prompts to authenticate with your OpenAI account.
+
+3. In **Settings → Agent Client**, leave the **API key field empty**.
+
+::: tip
+The ChatGPT app and Codex app use a different authentication system. Having ChatGPT running does **not** authenticate the plugin — you must log in via the CLI.
+:::
+
+## Verify Setup
+
+1. Click the robot icon in the ribbon or use the command palette: **"Open chat view"**
+2. Switch to Codex from the agent dropdown in the chat header
+3. Try sending a message to verify the connection
+
+Having issues? See [Troubleshooting](/help/troubleshooting).

package/docs/agent-setup/custom-agents.md

@@ -0,0 +1,67 @@
+# Custom Agents Setup
+
+You can use any agent that implements the [Agent Client Protocol (ACP)](https://agentclientprotocol.com/get-started/agents).
+
+## Install and Configure
+
+1. Install your ACP-compatible agent (e.g., [OpenCode](https://github.com/anomalyco/opencode), [Qwen Code](https://github.com/QwenLM/qwen-code), [Kiro](https://kiro.dev/)).
+
+2. Open **Settings → Agent Client** and scroll to **Custom Agents** section.
+
+3. Click **Add custom agent**.
+
+4. Configure the agent:
+   - **Agent ID**: Unique identifier (e.g., `my-agent`)
+   - **Display name**: Name shown in menus (e.g., `My Agent`)
+   - **Path**: Command name or absolute path to the agent executable. The command name alone (e.g., `opencode`) works in many cases. If the agent is not found automatically, set the full path, or click **Auto-detect**.
+   - **Arguments**: Command-line arguments, one per line (if required)
+   - **Environment variables**: `KEY=VALUE` pairs, one per line (if required)
+
+## Configuration Examples
+
+### OpenCode
+
+| Field | Value |
+|-------|-------|
+| **Agent ID** | `opencode` |
+| **Display name** | `OpenCode` |
+| **Path** | `opencode` |
+| **Arguments** | `acp` |
+| **Environment variables** | (optional) |
+
+### Qwen Code
+
+| Field | Value |
+|-------|-------|
+| **Agent ID** | `qwen-code` |
+| **Display name** | `Qwen Code` |
+| **Path** | `qwen` |
+| **Arguments** | `--experimental-acp` |
+| **Environment variables** | (optional) |
+
+### Kiro
+
+| Field | Value |
+|-------|-------|
+| **Agent ID** | `kiro-cli` |
+| **Display name** | `Kiro` |
+| **Path** | `kiro-cli` |
+| **Arguments** | `acp` |
+| **Environment variables** | (optional) |
+
+## Authentication
+
+Authentication depends on the specific agent. Common patterns:
+
+- **API Key**: Add to **Environment variables** (e.g., `MY_API_KEY=xxx`)
+- **Account Login**: Run the agent's CLI to authenticate, then leave environment variables empty
+
+Refer to your agent's documentation for specific authentication instructions.
+
+## Verify Setup
+
+1. Click the robot icon in the ribbon or use the command palette: **"Open chat view"**
+2. Select your custom agent from the agent dropdown in the chat header
+3. Try sending a message to verify the connection
+
+Having issues? See [Troubleshooting](/help/troubleshooting).

package/docs/agent-setup/gemini-cli.md

@@ -0,0 +1,99 @@
+# Gemini CLI Setup
+
+::: warning Gemini CLI is being discontinued
+Google is retiring account login for Gemini CLI (Pro / Ultra / free tiers) on **June 18, 2026**. Google states Gemini CLI stays accessible via a **paid** Gemini API key. See **[Gemini CLI Discontinuation & Migration](/announcements/gemini-cli-deprecation)**.
+:::
+
+Gemini CLI is Google's AI assistant. You can authenticate using your **Google account**, an **API key**, or **Vertex AI**.
+
+## Install and Configure
+
+Open a terminal (Terminal on macOS/Linux, PowerShell on Windows) and run the following commands.
+
+1. Install Gemini CLI:
+
+```bash
+npm install -g @google/gemini-cli
+```
+
+2. Find the installation path:
+
+::: code-group
+
+```bash [macOS/Linux]
+which gemini
+# Example output: /usr/local/bin/gemini
+```
+
+```cmd [Windows]
+where.exe gemini
+# Example output: C:\Users\Username\AppData\Roaming\npm\gemini.cmd
+```
+
+:::
+
+3. Open **Settings → Agent Client**. The default command (`gemini`) works in many cases. If the agent is not found automatically, set the **Gemini CLI path** to the path found above, or click **Auto-detect**.
+
+4. Ensure **Arguments** contains `--experimental-acp` (this is set by default).
+
+## Authentication
+
+Choose one of the following methods:
+
+### Option A: Google Account Login (OAuth)
+
+If you have a Google account and prefer not to use an API key, you can log in directly.
+
+::: warning
+Account login is being discontinued for Pro / Ultra / free tiers on June 18, 2026. See [Gemini CLI Discontinuation & Migration](/announcements/gemini-cli-deprecation).
+:::
+
+1. Run Gemini CLI in your terminal and choose "Login with Google":
+
+```bash
+gemini
+```
+
+2. Follow the browser authentication flow.
+
+3. In **Settings → Agent Client**, leave the **API key field empty**.
+
+::: tip
+If you have a Gemini Code Assist License from your organization, add `GOOGLE_CLOUD_PROJECT=YOUR_PROJECT_ID` in the **Environment variables** field.
+:::
+
+### Option B: Gemini API Key
+
+If you prefer to use an API key for authentication:
+
+1. Get your API key from [Google AI Studio](https://aistudio.google.com/apikey)
+2. Enter the API key in **Settings → Agent Client → Gemini CLI → API key**
+
+::: tip If the key isn't picked up
+If authentication still fails, set the key on the Gemini CLI side instead: run `gemini` in your terminal, type `/auth`, choose **Use Gemini API Key**, and paste your key (it's stored in your system keychain). Then leave the **API key field empty** in Agent Client.
+:::
+
+### Option C: Vertex AI
+
+If you are using Vertex AI for enterprise workloads:
+
+1. In **Settings → Agent Client → Gemini CLI → Environment variables**, add:
+
+```
+GOOGLE_API_KEY=YOUR_API_KEY
+GOOGLE_GENAI_USE_VERTEXAI=true
+```
+
+2. Leave the **API key field empty** (use Environment variables instead).
+
+::: tip If it isn't picked up
+If authentication still fails, configure Vertex AI on the Gemini CLI side instead: run `gemini` in your terminal, type `/auth`, and choose **Vertex AI**.
+:::
+
+## Verify Setup
+
+1. Click the robot icon in the ribbon or use the command palette: **"Open chat view"**
+2. Switch to Gemini CLI from the agent dropdown in the chat header
+3. Try sending a message to verify the connection
+
+Having issues? See [Troubleshooting](/help/troubleshooting).

package/docs/agent-setup/index.md

@@ -0,0 +1,34 @@
+# Agent Setup Overview
+
+Agent Client supports multiple AI agents through the [Agent Client Protocol (ACP)](https://github.com/agentclientprotocol/agent-client-protocol). This section covers how to set up each supported agent.
+
+## Supported Agents
+
+| Agent | Provider | Package |
+|-------|----------|---------|
+| [Claude Code](./claude-code) | Anthropic | `@agentclientprotocol/claude-agent-acp` |
+| [Codex](./codex) | OpenAI | `@zed-industries/codex-acp` |
+| [Gemini CLI](./gemini-cli) | Google | `@google/gemini-cli` |
+| [Custom Agents](./custom-agents) | Various | Any ACP-compatible agent |
+
+## Common Setup Steps
+
+All agents follow a similar setup pattern:
+
+1. **Install the agent package** via npm
+2. **Set up authentication** (API key or account login)
+
+The plugin resolves bare command names through your login shell's PATH, so path configuration is often not needed. If the agent is not found automatically, use `which` (macOS/Linux) or `where.exe` (Windows) to find the path and configure it in Settings → Agent Client.
+
+## WSL Mode (Windows)
+
+For Windows users, we recommend using **WSL Mode** for better compatibility:
+
+1. Install [WSL](https://docs.microsoft.com/en-us/windows/wsl/install)
+2. Install Node.js and agents inside WSL
+3. Enable **WSL Mode** in Settings → Agent Client
+4. Use Linux-style paths (e.g., `/usr/local/bin/node`)
+
+## Switching Agents
+
+Once you have multiple agents configured, you can switch between them using the **⋮** menu in the chat header. To change the default agent for new chat views, go to **Settings → Agent Client → Default agent**.

package/docs/announcements/gemini-cli-deprecation.md

@@ -0,0 +1,73 @@
+# Gemini CLI Discontinuation & Migration
+
+::: warning Gemini CLI is being discontinued
+Google is retiring **account login for Gemini CLI (Pro / Ultra / free tiers) on June 18, 2026**. After that date, signing in with your Google account will stop working in Gemini CLI.
+
+According to Google, **Gemini CLI stays accessible via a paid Gemini API key**, so the reliable way to keep using Gemini in this plugin is a paid API key.
+:::
+
+## What's changing
+
+On **June 18, 2026**, Gemini CLI will **stop serving requests** for account-login tiers. This is part of Google's transition from Gemini CLI to the new **Antigravity CLI**, announced at Google I/O 2026.
+
+Account login (signing in with your Google account) is what stops — but Google also points to a way to keep using Gemini CLI:
+
+> "Gemini CLI will remain accessible via paid Gemini and Gemini Enterprise Agent Platform API keys."
+> — [Google Developers Blog](https://developers.googleblog.com/an-important-update-transitioning-gemini-cli-to-antigravity-cli/)
+
+Note that Google's wording specifies a **paid** API key. The blog does not state whether a **free-tier** API key will keep working with Gemini CLI after the cutoff, so free-tier continuation is **not guaranteed**. This plugin drives Gemini CLI with an API key, so a paid API key is the dependable path forward.
+
+## Who is affected
+
+| Tier / access | Still usable after June 18, 2026? |
+| --- | --- |
+| Free tier (account login) | ❌ No — account login stops |
+| Google AI Pro (account login) | ❌ No — account login stops |
+| Google AI Ultra (account login) | ❌ No — account login stops |
+| **Gemini CLI via a paid API key** | ✅ Yes — keeps working |
+
+::: tip Note
+Being on a paid plan (Pro or Ultra) does **not** exempt you — the cutoff is about the *account-login* method, not whether you pay. The reliable way to keep going is a **paid API key** (see below).
+:::
+
+## How to keep using Gemini in this plugin
+
+### Option A — Use a paid Gemini API key (recommended)
+
+1. Create an API key in [Google AI Studio](https://aistudio.google.com/apikey) and enable billing to move it to a [paid tier](https://ai.google.dev/gemini-api/docs/billing). This is the access Google explicitly says keeps working.
+2. Configure the key in this plugin by following the [Gemini CLI Setup](/agent-setup/gemini-cli) page (Option B: API key, or Option C: Vertex AI).
+
+::: tip About the free tier
+The Gemini API has a free tier, and as of writing models such as `gemini-2.5-pro` and `gemini-3.5-flash` show as free of charge on the [pricing page](https://ai.google.dev/gemini-api/docs/pricing). **However, Google's announcement only guarantees access via a *paid* API key** — it does not say a free-tier key will keep working with Gemini CLI after June 18, 2026. A free key *may* work, but it is not guaranteed; if you need reliability, use a paid key. Rate limits for any tier can be checked in the [AI Studio rate-limit dashboard](https://aistudio.google.com/rate-limit).
+:::
+
+::: tip Privacy note
+A paid API key also helps with privacy. Agent Client can send your **note contents** to the model as context, and on **paid** tiers Google does not use that data to improve its products: *"When you use Paid Services, Google doesn't use your prompts ... or responses to improve our products."* (If you ever fall back to a free-tier key, note that free/unpaid usage **can** be used for product improvement and human review.) Source: [Gemini API Additional Terms of Service](https://ai.google.dev/gemini-api/terms).
+:::
+
+### Option B — Switch to another ACP agent
+
+This plugin works with any agent that speaks the Agent Client Protocol (ACP). If you'd rather move off Gemini entirely:
+
+- [Claude Code](/agent-setup/claude-code)
+- [Codex](/agent-setup/codex)
+
+## What about Antigravity CLI?
+
+Google's replacement, **Antigravity CLI**, does **not currently support ACP** (there is no `--experimental-acp` mode), so this plugin **cannot connect to it as a built-in agent yet**. Antigravity CLI is also not open source.
+
+If Google adds ACP support — or if an adapter built on the **Antigravity SDK** emerges (similar to how `claude-agent-acp` is built on the Claude Agent SDK rather than wrapping the CLI) — built-in support will be reconsidered. For Google's own migration steps, see the [Antigravity CLI migration guide](https://antigravity.google/docs/gcli-migration).
+
+A community ACP bridge, [`agy-acp`](https://github.com/openabdev/openab/tree/main/agy-acp), does exist and can be added as a [custom agent](/agent-setup/custom-agents). It works as a minimal text chat only — streaming, tool calls, and diffs don't come through — so it's an escape hatch rather than a full replacement.
+
+## Questions?
+
+If anything is unclear or you run into trouble migrating, join the discussion on [GitHub](https://github.com/RAIT-09/obsidian-agent-client/discussions/300).
+
+## Sources
+
+- [An important update: Transitioning Gemini CLI to Antigravity CLI — Google Developers Blog](https://developers.googleblog.com/an-important-update-transitioning-gemini-cli-to-antigravity-cli/)
+- [Google I/O 2026 developer highlights — Google blog](https://blog.google/intl/ja-jp/company-news/technology/google-io-2026-developer-highlights/)
+- [Gemini Developer API pricing](https://ai.google.dev/gemini-api/docs/pricing)
+- [Gemini API Additional Terms of Service](https://ai.google.dev/gemini-api/terms)
+- [Gemini API rate limits](https://ai.google.dev/gemini-api/docs/rate-limits)

package/docs/getting-started/index.md

@@ -0,0 +1,78 @@
+# Installation
+
+## Install the Plugin
+
+### From Community Plugins (Recommended)
+
+1. In Obsidian, open **Settings → Community Plugins → Browse**
+2. Search for **"Agent Client"**
+3. Click **Install**, then **Enable**
+
+<p align="center">
+  <a href="obsidian://show-plugin?id=agent-client" style="display:inline-block;padding:10px 24px;background:#7c3aed;color:white;border-radius:8px;text-decoration:none;font-weight:600;">Add to Obsidian</a>
+</p>
+
+### Via BRAT (Pre-release Versions)
+
+To try pre-release versions before they are published to Community Plugins:
+
+1. Install the [BRAT](https://github.com/TfTHacker/obsidian42-brat) plugin from the Community Plugins browser
+2. In Obsidian settings, go to **Community Plugins → BRAT → Add Beta Plugin**
+3. Paste this repo URL:
+   ```
+   https://github.com/RAIT-09/obsidian-agent-client
+   ```
+4. BRAT will download the latest release and keep it auto-updated
+5. Enable **Agent Client** from the plugin list
+
+### Manual Installation
+
+1. Download the latest release files from [GitHub Releases](https://github.com/RAIT-09/obsidian-agent-client/releases):
+   - `main.js`
+   - `manifest.json`
+   - `styles.css`
+2. Create the plugin folder: `VaultFolder/.obsidian/plugins/agent-client/`
+3. Place the downloaded files in this folder
+4. Enable the plugin in **Obsidian Settings → Community Plugins**
+
+## Prerequisites
+
+### Node.js
+
+::: tip Not always required
+Node.js is needed for npm-based agents like Claude Code, Codex, and Gemini CLI. If your agent is a standalone binary, you can skip this step.
+:::
+
+If you need Node.js:
+
+1. Download from [nodejs.org](https://nodejs.org/)
+2. Install the LTS version (recommended)
+
+### Find Your Node.js Path
+
+If auto-detect doesn't find Node.js, you can locate it manually. Open a terminal (Terminal on macOS/Linux, PowerShell on Windows) and run:
+
+::: code-group
+
+```bash [macOS/Linux]
+which node
+# Example output: /usr/local/bin/node
+```
+
+```cmd [Windows]
+where.exe node
+# Example output: C:\Program Files\nodejs\node.exe
+```
+
+:::
+
+### Configure Node.js Path
+
+In most cases, the plugin automatically finds Node.js through your login shell's PATH, so no configuration is needed. If Node.js is not detected automatically:
+
+1. Open **Settings → Agent Client**
+2. Click the **Auto-detect** button next to the **Node.js path** field, or enter the path manually
+
+## Next Steps
+
+Continue to [Quick Start](./quick-start) to set up your first agent and start chatting!

package/docs/getting-started/quick-start.md

@@ -0,0 +1,38 @@
+# Quick Start
+
+This guide will get you chatting with an AI agent in just a few minutes.
+
+## Step 1: Choose Your Agent
+
+Agent Client supports multiple AI agents. Choose one to start:
+
+| Agent | Provider | Integration |
+|-------|----------|-------------|
+| **[Claude Code](/agent-setup/claude-code)** | Anthropic | via [ACP adapter](https://github.com/agentclientprotocol/claude-agent-acp) |
+| **[Codex](/agent-setup/codex)** | OpenAI | via [Zed's adapter](https://github.com/zed-industries/codex-acp) |
+| **[Gemini CLI](/agent-setup/gemini-cli)** | Google | with `--experimental-acp` option |
+| **[Custom](/agent-setup/custom-agents)** | Various | [Any ACP-compatible agent](https://agentclientprotocol.com/overview/agents) (e.g., OpenCode, Qwen Code, Kiro) |
+
+## Step 2: Install and Configure the Agent
+
+Follow the setup guide for your chosen agent:
+
+- [Claude Code Setup](/agent-setup/claude-code)
+- [Codex Setup](/agent-setup/codex)
+- [Gemini CLI Setup](/agent-setup/gemini-cli)
+- [Custom Agents](/agent-setup/custom-agents)
+
+Each guide covers installation, path configuration, and authentication.
+
+## Step 3: Start Chatting
+
+1. Click the **robot icon** in the left ribbon, or
+2. Open the command palette (`Cmd/Ctrl + P`) and search for **"Open chat view"**
+
+The chat panel opens in the right sidebar. Type a message and press Enter!
+
+## What's Next?
+
+- Learn about [Note Mentions](/usage/mentions) to reference your notes in conversations
+- Explore [Slash Commands](/usage/slash-commands) for quick actions
+- Set up additional agents in [Agent Setup](/agent-setup/)