docs: browser extensions, terminal link menu, and a wording trim

apps/docs/content/docs/add-agent-support.mdx

@@ -5,7 +5,7 @@ description: For agent makers - how your coding agent runs in Paradev today, and
 
 ## Overview
 
-If you build a terminal coding agent, it already runs in Paradev: users can add any command as a custom agent, and it gets its own workspace, Git worktree, and terminal. First-class support adds status indicators, notifications, session resume, and usage tracking. This page explains what each one needs from your agent, so Paradev users get the best experience with it.
+A terminal coding agent already runs in Paradev: users can add any command as a custom agent, and it gets its own workspace, Git worktree, and terminal. First-class support adds status indicators, notifications, session resume, and usage tracking. This page lists what each needs from your agent.
 
 ## Works Today: Custom Agents
 
@@ -15,7 +15,7 @@ Users add your agent in **Settings → Agents → Add agent → Custom agent…*
 - **Prompt-only args** and **Prompt transport** - how Paradev passes a prompt: as an argument or on stdin
 - **Resume args** and **Fork args** - how to reopen or clone a session by its id, if your agent supports it
 
-To make this easy, document a single command that starts your agent, a way to pass the first prompt as an argument or on stdin, and, if you have them, resume and fork flags that take a session id.
+Document one command that starts your agent, a way to pass the first prompt as an argument or on stdin, and, if you have them, resume and fork flags that take a session id.
 
 Inside a Paradev terminal, your agent can read these environment variables:
 
@@ -28,7 +28,7 @@ Inside a Paradev terminal, your agent can read these environment variables:
 
 ## First-Class Support
 
-Built-in agents get more than a terminal. Each feature below has a clear requirement:
+Built-in agents get more than a terminal. Each feature has a requirement:
 
 | Feature | What your agent needs |
 |---|---|
@@ -41,6 +41,4 @@ Built-in agents get more than a terminal. Each feature below has a clear require
 
 Built-in agent support lives in the Paradev repository on GitHub. To request it, [open an issue](https://github.com/paradevai/paradev/issues) with your agent's install command, its hook and session-id documentation, and where it writes session logs. We review pull requests that add an agent too.
 
-## What Users See
-
-With first-class support, your agent appears in Paradev's agent launcher with its icon, shows a live status in the sidebar, sends a notification when it needs input or finishes, and comes back with its session after a restart. See [Supported Agents](/supported-agents) for where each built-in agent stands today.
+See [Supported Agents](/supported-agents) for where each built-in agent stands today.

apps/docs/content/docs/agent-integration.mdx

@@ -5,7 +5,7 @@ description: Use Claude Code and other AI agents
 
 ## Overview
 
-Run AI coding agents in separate workspaces. Each agent gets its own Git worktree and branch, so agents don't edit the same working files.
+Run AI coding agents in separate workspaces. Each workspace has its own Git worktree and branch, so agents don't edit the same files.
 
 ![AI Agents](/images/agents.png)
 
@@ -36,19 +36,15 @@ Run AI coding agents in separate workspaces. Each agent gets its own Git worktre
 
 ## Chat
 
-Paradev includes a built-in chat panel for AI interactions without leaving the app.
+Paradev has a built-in chat panel.
 
-- Use the model picker to choose supported models and connect providers
-- Chat session titles are generated automatically from the conversation
-- Use the thinking level control to adjust reasoning effort for the selected model
+- Use the model picker to choose supported models and connect providers.
+- Session titles are generated from the conversation.
+- Use the thinking level control to set reasoning effort for the selected model.
 
 ## Models and Reasoning Effort
 
-The model pickers include **GPT-6 Astra**, **GPT-6 Sol**, and **GPT-6 Luna**,
-and **Opus 5.5** and **Sonnet 5.5**, wherever the agent supports them: Claude
-Code (Claude models), Codex (GPT models), Copilot, OpenCode, Oh My Pi, and Paradev
-Chat. The Codex picker also has **GPT-6.1 Sol** and **GPT-6.1 Terra**; the
-retired GPT-5.5, GPT-5.4, and GPT-5.3 Codex are gone. Available reasoning levels depend on the agent and model; see the [launch options](/cli/cli-reference#agent-effort-levels).
+The model pickers include **GPT-6 Astra**, **GPT-6 Sol**, **GPT-6 Luna**, **Opus 5.5**, and **Sonnet 5.5**, wherever the agent supports them: Claude Code (Claude models), Codex (GPT models), Copilot, OpenCode, Oh My Pi, and Paradev Chat. The Codex picker also has **GPT-6.1 Sol** and **GPT-6.1 Terra**; GPT-5.5, GPT-5.4, and GPT-5.3 Codex are retired. Available reasoning levels depend on the agent and model; see the [launch options](/cli/cli-reference#agent-effort-levels).
 
 ## Agent Settings
 
@@ -90,23 +86,23 @@ Resume an interrupted session, fork a session, or hand its work to another agent
 
 ## Parallel Agents
 
-Create multiple workspaces and run an agent in each. Compare results and merge the best solution.
+Run an agent in each of several workspaces, compare results, and merge the best one.
 
-## Inline Prompts
+## Sandbox Access
 
-Agents can pause and ask questions, request tool approval, and ask for plan review directly in the chat pane.
+When an agent needs files outside its workspace, it asks for permission in chat. You approve or deny inline.
 
-## Sandbox Access
+## Inline Prompts
 
-When an agent needs access to files outside its workspace, it will prompt for permission in chat so you can approve or deny the request inline.
+Prompts, approvals, and plan reviews appear in the chat pane.
 
 ## Notifications
 
-Get notified when agents finish. Configure sounds in Settings.
+You are notified when agents finish. Configure sounds in Settings. See [Agent Status & Notifications](/agent-status).
 
 ## Agent Skills
 
-Teach any agent to drive Paradev itself with the official [skills](https://github.com/paradevai/skills), one for the `superset` CLI and one for the [MCP server](/mcp-server):
+Teach an agent to drive Paradev with the official [skills](https://github.com/paradevai/skills), one for the `superset` CLI and one for the [MCP server](/mcp-server):
 
 ```bash
 npx skills add paradevai/skills

apps/docs/content/docs/agent-sessions.mdx

@@ -3,7 +3,7 @@ title: Agent Sessions & Forking
 description: Resume, fork, and hand off agent sessions
 ---
 
-An agent session is one agent conversation running in one terminal. Paradev can bring a session back after its terminal dies, copy it into a new session of the same agent, or pass its recent context to a different agent.
+An agent session is one agent conversation in one terminal. Paradev can resume a session after its terminal dies, fork it into a new session of the same agent, or pass its recent context to a different agent.
 
 ## Resume Interrupted Sessions
 

apps/docs/content/docs/agent-status.mdx

@@ -7,23 +7,23 @@ description: How Paradev knows what your agents are doing, and how it tells you
 
 Each workspace in the sidebar shows an activity strip:
 
-- **Running agents and open ports** appear inline on the workspace item
-- **Running and PR state persist across restarts**
-- Workspaces with live work are visually emphasized so in-progress items stand out
-- Agent chips stay hidden unless multiple agents are running in that workspace
+- Running agents and open ports appear inline on the workspace item.
+- Running and PR state persist across restarts.
+- Workspaces with live work are emphasized.
+- Agent chips show only when several agents run in that workspace.
 
 If a workspace shows as running after you've stopped its agent (for example after force-quitting the agent), use **Clear Status** on the workspace to reset the indicator.
 
 ## How Status Is Detected
 
-Paradev installs lifecycle hooks and command wrappers into each agent's config (the entries in `~/.paradev/bin` and files like `superset-hooks`). Those hooks report when an agent starts, finishes, or is waiting for input. That's why status only works for agents launched through Paradev, and why removing the hooks disables status and notifications.
+Paradev installs lifecycle hooks and command wrappers into each agent's config (the entries in `~/.paradev/bin` and files like `superset-hooks`). Those hooks report when an agent starts, finishes, or is waiting for input. Status therefore works only for agents launched through Paradev, and removing the hooks disables status and notifications.
 
 The hooks no-op outside Paradev terminals (they check for a Paradev-supplied env var). To remove them from an agent's config entirely, turn off the agent's **Paradev hooks** toggle in **Settings → Agents**.
 
 ## Notifications
 
-- **Finished / waiting for input**: get notified when an agent finishes or pauses for your input. Configure sounds in Settings.
-- **Support varies by agent.** Claude Code reports both finished and waiting states. Some agents don't emit a waiting-for-input signal yet, so you'll only get completion notifications for them.
+- **Finished / waiting for input**: you are notified when an agent finishes or pauses for your input. Configure sounds in Settings.
+- **Support varies by agent.** Claude Code reports both finished and waiting states. Some agents send no waiting-for-input signal, so you get completion notifications only. See [Supported Agents](/supported-agents).
 - **Inline prompts**: agents can pause to ask questions, request tool approval, or ask for plan review directly in the chat pane, and the workspace is flagged as needing attention.
 
 ## Dock Badge

apps/docs/content/docs/agents/claude-code.mdx

@@ -28,4 +28,4 @@ This turns off Claude Code's approval prompts, so the agent can edit files and r
 
 ## Run Several Sessions
 
-Press **⌘N** for each task. Each workspace gets its own worktree, so sessions never edit the same files. To give the same task to Claude Code and another agent and keep the best result, see [Race Agents on One Task](/recipes/race-agents).
+Press **⌘N** for each task. Each workspace has its own worktree, so sessions don't edit the same files. To give the same task to Claude Code and another agent and keep the best result, see [Race Agents on One Task](/recipes/race-agents).

apps/docs/content/docs/agents/codex.mdx

@@ -28,4 +28,4 @@ This turns off Codex's approval prompts and sandbox, so the agent can edit files
 
 ## Run Several Sessions
 
-Press **⌘N** for each task. Each workspace gets its own worktree, so sessions never edit the same files. To give the same task to Codex and another agent and keep the best result, see [Race Agents on One Task](/recipes/race-agents).
+Press **⌘N** for each task. Each workspace has its own worktree, so sessions don't edit the same files. To give the same task to Codex and another agent and keep the best result, see [Race Agents on One Task](/recipes/race-agents).