# AI use cases for Grok App (RongleCat/grok-app) Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Use cases: 5 | Page: https://dedupedai.com/use-cases/ronglecat-grok-app/ Grok App is an open-source desktop workbench that gives a GUI to the local Grok Build CLI (grok agent stdio), covering sessions, projects, files, media, and automations. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Run a local Grok agent on a project from a desktop app: https://dedupedai.com/use-cases/ronglecat-grok-app/run-local-grok-agent-on-project/ - Run parallel agent tasks in Git worktrees and review diffs: https://dedupedai.com/use-cases/ronglecat-grok-app/review-agent-edits-with-git-worktrees/ - Check and steer your desktop agent from a phone: https://dedupedai.com/use-cases/ronglecat-grok-app/steer-local-agent-from-phone/ - Script agent sessions through the loopback REST API: https://dedupedai.com/use-cases/ronglecat-grok-app/script-agent-sessions-via-local-api/ - Schedule recurring agent tasks with natural-language prompts: https://dedupedai.com/use-cases/ronglecat-grok-app/schedule-recurring-agent-tasks/ ## How to run a local Grok agent on a project from a desktop app with Grok App Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Difficulty: Beginner | Time: 30 minutes | Basis: From the README Install the desktop workbench, connect your Grok Build CLI, and work on a project folder with approval prompts. **Who it is for:** A developer who wants a visual workbench instead of a terminal for running the Grok Build agent on a local codebase. **The problem:** Running the agent only in a terminal makes it hard to track several tasks, review edits, and keep project context organized. **The outcome:** A trusted project open in the desktop app with an agent session running under Ask permission mode. ### Steps 1. Download the installer for your platform from Grok App.com or the GitHub Releases page. 2. Launch Grok App and let the setup wizard verify the Grok Build CLI, or run `grok update` in a terminal if the CLI is already installed. 3. On the account step, choose Use existing CLI sign-in if your local grok CLI is already authenticated. 4. Add your project by selecting its working directory and trusting it. 5. Choose Ask as the permission mode and start a session with a first prompt. ### Prompt to paste into your agent ```text Using the Grok App repo (https://github.com/RongleCat/grok-app), set up the desktop workbench and open the project at as a trusted workspace in Ask permission mode. Expected output: a running agent session tied to that project. Verify that the grok CLI version is 0.2.112 or newer before starting. ``` ### Caveat Grok App is not an official xAI product and needs a working local grok CLI for full agent capabilities. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/ronglecat-grok-app/run-local-grok-agent-on-project/ ## How to run parallel agent tasks in Git worktrees and review diffs with Grok App Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Switch between linked Git worktrees and accept or revert agent changes file by file. **Who it is for:** A developer maintaining a Git repository who wants several agent tasks running at once without mixing their changes. **The problem:** Parallel agent edits in one checkout are hard to separate and review before committing. **The outcome:** Separate sessions working in different worktrees, with each set of changes reviewed through the Git diff view before acceptance. ### Steps 1. Create or confirm linked Git worktrees for the repository you want to work on (inferred: `git worktree add `). (inferred, check the README) 2. Open the repository as a project in Grok App so the linked worktrees are discovered. 3. Start one session per worktree and give each agent a separate task. 4. Use the Kanban board to track which sessions need input, are working, or are done. 5. Open the visual Git Diff review and accept, reject, or revert changes for single files or in batches. ### Prompt to paste into your agent ```text Using the Grok App repo (https://github.com/RongleCat/grok-app), open as a project and start one agent session per linked Git worktree. Expected output: separate sessions with their own changes visible in the Git diff review. Verify that each session shows the worktree it was assigned to before accepting any edits. ``` ### Caveat The worktree discovery depends on the repository already having linked worktrees, and the README does not describe how to create them inside the app. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/ronglecat-grok-app/review-agent-edits-with-git-worktrees/ ## How to check and steer your desktop agent from a phone with Grok App Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Use the token-gated mobile web mirror or an IM bridge to follow and redirect a running session. **Who it is for:** A developer who starts long agent tasks at a desk and wants to check progress or redirect them while away. **The problem:** Agent tasks keep running on the desktop, and there is no way to check or adjust them without returning to the machine. **The outcome:** A phone view of the running session where you can monitor progress and send a steering message. ### Steps 1. Open the desktop app and enable the Mobile Web Mirror from its settings. 2. Note the access token that the mirror requires. 3. Optionally set up one of the IM channel bridges such as Telegram or Slack to reach the same agent. 4. On the phone, open the mirror address in a mobile browser and enter the token. 5. Send a follow-up prompt or steering message to the active session, or use `/p` or `/r` in an IM bridge to switch project or resume a session. ### Prompt to paste into your agent ```text Using the Grok App repo (https://github.com/RongleCat/grok-app), enable the mobile web mirror for a running session in . Expected output: a token-gated mobile page that shows the session timeline. Verify that the token is required before any page content loads. ``` ### Caveat Remote access options such as Cloudflare Quick Tunnel are named in the README, and the setup details beyond that are not documented there. Risk, remote script: Remote messages can direct a local agent that runs tools on your machine, so keep the access token private and use Ask mode unless you fully trust the channel. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/ronglecat-grok-app/steer-local-agent-from-phone/ ## How to script agent sessions through the loopback REST API with Grok App Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Call the local session endpoints from scripts or CI to list sessions and send turns. **Who it is for:** A developer building a script or CI step that needs to start or continue agent work on a local machine. **The problem:** Starting agent work by hand in the GUI does not fit repeatable scripts or pipelines. **The outcome:** A script that lists existing sessions and posts a new turn to a chosen session. ### Steps 1. Launch Grok App and keep it running on the machine that hosts the sessions. 2. List the existing sessions with a GET request to the loopback endpoint `GET /v1/sessions`. 3. Pick a session identifier from the response. 4. Send a new prompt with a POST request to `POST /v1/sessions/{id}/turns`, replacing {id} with the chosen identifier. 5. Check the session in the desktop timeline to confirm the turn arrived. ### Prompt to paste into your agent ```text Using the Grok App repo (https://github.com/RongleCat/grok-app), call the loopback session API to list sessions and post one turn to session with the prompt . Expected output: a successful response and a new turn in that session's timeline. Verify the session identifier against the GET /v1/sessions response before posting. ``` ### Caveat The README names the two endpoints but does not document request or response bodies, port numbers, or authentication, so check the app for those details before relying on them. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/ronglecat-grok-app/script-agent-sessions-via-local-api/ ## How to schedule recurring agent tasks with natural-language prompts with Grok App Repo: [RongleCat/grok-app](https://github.com/RongleCat/grok-app) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Set up one-off or repeating agent jobs and review their execution logs. **Who it is for:** A developer who wants routine checks or reports produced by the agent on a schedule. **The problem:** Routine agent tasks like status checks need someone to start them manually each time. **The outcome:** An automation that runs on a schedule and leaves a visual execution log to review. ### Steps 1. Open the Scheduled Automations area in Grok App. 2. Describe the task in plain language and let the app create the prompt. 3. Choose whether the task is recurring or one-off and select the project it should run in. 4. Keep Ask permission mode unless you have decided the task can run unattended. 5. Open the visual execution log after the first run to check the output. ### Prompt to paste into your agent ```text Using the Grok App repo (https://github.com/RongleCat/grok-app), create a scheduled automation for project that runs the task on . Expected output: an automation entry with an execution log after its first run. Verify the permission mode before the first scheduled run. ``` ### Caveat The README does not describe the schedule syntax or limits, so verify how recurrence is set in the app before relying on it. Risk, remote script: A scheduled task can run agent tools without a person watching, so use Ask mode for anything that changes files or runs commands. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/ronglecat-grok-app/schedule-recurring-agent-tasks/