# AI use cases for Herdr Web UI (devswha/herdr-web-ui) Repo: [devswha/herdr-web-ui](https://github.com/devswha/herdr-web-ui) | Use cases: 4 | Page: https://dedupedai.com/use-cases/devswha-herdr-web-ui/ Ovis (repo devswha/herdr-web-ui) is a browser and phone client for herdr that shows the chat and live terminal of each agent pane on your computer. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Reply to Claude Code or Codex from your phone: https://dedupedai.com/use-cases/devswha-herdr-web-ui/reply-to-agents-from-phone/ - Get push alerts when an agent needs input: https://dedupedai.com/use-cases/devswha-herdr-web-ui/get-push-alerts-when-agent-needs-input/ - Review repository changes from the file browser: https://dedupedai.com/use-cases/devswha-herdr-web-ui/review-repo-changes-without-editing/ - Dictate prompts to an agent by voice: https://dedupedai.com/use-cases/devswha-herdr-web-ui/dictate-prompts-to-an-agent/ ## How to reply to Claude Code or Codex from your phone with Herdr Web UI Repo: [devswha/herdr-web-ui](https://github.com/devswha/herdr-web-ui) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Install the client, add it to your phone, and answer an agent's question from chat. **Who it is for:** A developer whose Claude Code or Codex session on a home PC is waiting for input while they are away from the desk. **The problem:** The agent stops at a question or approval, and the developer cannot reach the terminal running on the PC. **The outcome:** The same agent pane appears in the phone browser as a chat, where approvals and questions can be tapped, and the live terminal is one tap away. ### Steps 1. Run `curl -fsSL https://herdrweb.dev/install.sh | sh` on the PC that runs herdr. 2. Open `http://127.0.0.1:7317` on the PC, or use the address the installer prints. 3. On your phone, scan the installer's QR code, or open Phone setup in herdr. 4. Add the app to your phone's home screen from the browser menu. 5. Open a Claude Code or Codex pane in the app and reply in the chat, or switch to Terminal for requests without card support. ### Prompt to paste into your agent ```text Use the repo devswha/herdr-web-ui. Install it on this PC with the README install command, then open the agent pane named in the browser client and list any pending approval or question shown as a card. Expected output: the pane name, its state, and the pending items, or a statement that none are pending. Verify that the address responds with HTTP 200 before reporting. ``` ### Caveat Reaching the PC from a phone depends on Tailscale or another route you set up, and the README warns against exposing the port through a public proxy without a token. Risk, remote script: The install command pipes a remote script into sh, so read it first if your policy requires review before running it. 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/devswha-herdr-web-ui/reply-to-agents-from-phone/ ## How to get push alerts when an agent needs input with Herdr Web UI Repo: [devswha/herdr-web-ui](https://github.com/devswha/herdr-web-ui) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Receive a phone alert when an agent asks a question or finishes, even with the app closed. **Who it is for:** A developer who starts long agent tasks in herdr and wants to leave the desk without checking the terminal. **The problem:** An agent waits for a reply or completes a task unnoticed while the developer works on something else. **The outcome:** The phone receives a push alert when an agent needs input or finishes, and a tap opens the pending question. ### Steps 1. Install the client with `curl -fsSL https://herdrweb.dev/install.sh | sh` on the PC that runs herdr. 2. Open the app on a secure address such as `http://127.0.0.1:7317` or an HTTPS address, since installation and push need a secure context. 3. Allow notifications for the installed app on your phone when the browser asks. 4. Start an agent task in a herdr pane and leave the app closed on the phone. 5. When the alert arrives, tap it to open the pending question in the chat. ### Prompt to paste into your agent ```text Use the repo devswha/herdr-web-ui. Confirm that push alerts are set up for this PC, then report the address the client uses and whether it is a secure context. Expected output: a short status with the address and a yes or no for push readiness. Verify the address uses HTTPS or localhost before saying push is ready. ``` ### Caveat Push alerts need a secure context such as HTTPS or localhost, and the README does not describe how to set up HTTPS for a phone, so check your own route before relying on alerts. 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/devswha-herdr-web-ui/get-push-alerts-when-agent-needs-input/ ## How to review repository changes from the file browser with Herdr Web UI Repo: [devswha/herdr-web-ui](https://github.com/devswha/herdr-web-ui) | Difficulty: Beginner | Time: 15 minutes | Basis: From the README Read each file's diff for working-tree, staged or commit comparisons without changing the repo. **Who it is for:** A developer or reviewer who wants to inspect what an agent changed in a repository from the browser. **The problem:** An agent has edited several files, and the reviewer wants to read the diffs before accepting the work. **The outcome:** The reviewer reads each changed file's diff in the browser, and the repository stays unchanged. ### Steps 1. Install the client with `curl -fsSL https://herdrweb.dev/install.sh | sh`, or with `herdr plugin install devswha/herdr-web-ui` if herdr is already set up. 2. Open `http://127.0.0.1:7317` in the browser. 3. Open the repository's folder in the file browser. 4. Choose a working-tree, staged or commit comparison. 5. Open each changed file to read its diff. ### Prompt to paste into your agent ```text Use the repo devswha/herdr-web-ui. Open the repository at in the file browser and list the files changed in the working tree with their diff status. Expected output: a list of changed file paths and their status. Verify that no files in were modified before reporting. ``` ### Caveat The README says diffs are read without changing the repository, but comparisons only cover the repositories that herdr and the client can reach on the PC running 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/devswha-herdr-web-ui/review-repo-changes-without-editing/ ## How to dictate prompts to an agent by voice with Herdr Web UI Repo: [devswha/herdr-web-ui](https://github.com/devswha/herdr-web-ui) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Speak a message in mixed Korean and English, review the text, then send it to the agent. **Who it is for:** A developer who wants to give an agent a longer instruction by voice from a phone instead of typing it. **The problem:** Typing a long prompt on a phone keyboard is slow, and the developer switches languages mid-sentence. **The outcome:** The spoken words appear as text in the chat or terminal line for review, and the agent receives them only after the developer sends them. ### Steps 1. Install the client with `curl -fsSL https://herdrweb.dev/install.sh | sh` on the PC that runs herdr. 2. Open the agent pane in the phone browser and tap the dictation control in the chat. 3. Speak the message, mixing Korean and English if needed. 4. Review the transcribed words and correct them before sending. 5. Send the message to the agent, or queue it for the next turn. ### Prompt to paste into your agent ```text Use the repo devswha/herdr-web-ui. Check whether dictation is configured on this PC by reporting whether an OpenAI key is set or the browser speech recognition is in use. Expected output: a one-line status naming the transcription source. Verify the source by checking the app settings before reporting. ``` ### Caveat Voice input sends audio to the configured provider, or to the browser's speech service, so use it only where that data flow is acceptable. 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/devswha-herdr-web-ui/dictate-prompts-to-an-agent/