# AI use cases for Pocket Strike (AbuZar-Ansarii/PocketStrike) Repo: [AbuZar-Ansarii/PocketStrike](https://github.com/AbuZar-Ansarii/PocketStrike) | Use cases: 4 | Page: https://dedupedai.com/use-cases/abuzar-ansarii-pocketstrike/ PocketStrike is an open-source Flutter app that turns an Android phone into an AI workstation with on-device GGUF models, cloud LLM providers, a ReAct agent, MCP servers, memory files, and image generation. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Build and verify the Pocket Strike release APK: https://dedupedai.com/use-cases/abuzar-ansarii-pocketstrike/build-and-verify-pocketstrike-apk/ - Run a private GGUF model on your phone with no cloud: https://dedupedai.com/use-cases/abuzar-ansarii-pocketstrike/run-offline-gguf-model-on-phone/ - Generate images from a chat with /image commands: https://dedupedai.com/use-cases/abuzar-ansarii-pocketstrike/generate-images-from-chat-prompts/ - Connect MCP servers to a ReAct agent on your phone: https://dedupedai.com/use-cases/abuzar-ansarii-pocketstrike/connect-mcp-servers-to-phone-agent/ ## How to build and verify the Pocket Strike release APK with Pocket Strike Repo: [AbuZar-Ansarii/PocketStrike](https://github.com/AbuZar-Ansarii/PocketStrike) | Difficulty: Intermediate | Time: 1.5 hours | Basis: From the README Clone the repo, run analysis and tests, and produce a release APK from source. **Who it is for:** A Flutter or Android developer who wants to inspect or ship their own build of Pocket Strike instead of trusting a prebuilt binary. **The problem:** Installing an app that handles API keys, files, and agent actions is easier to trust when you can build it yourself and check the code passes analysis and tests. **The outcome:** A release APK at build/app/outputs/flutter-apk/app-release.apk, with static analysis and the test suite run first. ### Steps 1. Run `git clone https://github.com/your-username/PocketStrike.git` (replace your-username with the actual GitHub owner if the README placeholder is outdated), then `cd PocketStrike`. 2. Run `flutter pub get` to install dependencies. 3. Run `flutter analyze` and confirm it reports no warnings or errors, then run `flutter test`. 4. Run `flutter build apk --release`. 5. Install build/app/outputs/flutter-apk/app-release.apk on your Android device. ### Prompt to paste into your agent ```text Clone the PocketStrike repo at and run flutter pub get, flutter analyze, flutter test, and flutter build apk --release. Report the analyzer result, the test result, and the path of the generated APK. Verify that build/app/outputs/flutter-apk/app-release.apk exists before reporting success. ``` ### Caveat The README's git clone URL uses a your-username placeholder, so confirm the real repository owner before cloning. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/abuzar-ansarii-pocketstrike/build-and-verify-pocketstrike-apk/ ## How to run a private GGUF model on your phone with no cloud with Pocket Strike Repo: [AbuZar-Ansarii/PocketStrike](https://github.com/AbuZar-Ansarii/PocketStrike) | Difficulty: Intermediate | Time: 1.5 hours | Basis: From the README Download a small quantized model and chat with it fully on-device. **Who it is for:** A person who wants to use an AI assistant on an Android phone without sending prompts or data to any cloud service. **The problem:** Cloud chat apps send every prompt off the device. You want answers that stay on your phone. **The outcome:** A chat session answered by a local model, with live tokens-per-second shown in the response. ### Steps 1. Build and install Pocket Strike on your Android device using the Flutter commands in the README, starting with `flutter pub get` and `flutter run -d` (add your device ID after -d). 2. Open the local models area and pick a preset such as Qwen2.5-0.5B-Instruct, which lists about 450 MB RAM footprint in the README table. 3. Download the model in the app, then tap Load into RAM. 4. Start a new conversation and send a prompt. 5. Check the bottom-right of the AI response for the tok/s speed badge. ### Prompt to paste into your agent ```text In the PocketStrike repo, help the user install the app on a connected Android device with flutter run -d , then download the Qwen2.5-0.5B-Instruct preset and load it into RAM. Expected output: a chat reply generated locally with a tok/s badge shown. Verify that the reply appears with the model loaded and no cloud provider selected. ``` ### Caveat Model downloads come from HuggingFace and need storage and RAM that match the model size in the README table, so check the footprint before loading a larger model. Risk, reverse engineering: The uncensored presets remove alignment refusals, so treat their output as unfiltered and check it before acting on it. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/abuzar-ansarii-pocketstrike/run-offline-gguf-model-on-phone/ ## How to generate images from a chat with /image commands with Pocket Strike Repo: [AbuZar-Ansarii/PocketStrike](https://github.com/AbuZar-Ansarii/PocketStrike) | Difficulty: Intermediate | Time: 1.5 hours | Basis: From the README Create images in any conversation by typing /image with an aspect ratio. **Who it is for:** A creator or hobbyist who wants quick image concepts for wallpapers, posts, or drafts while working on a phone. **The problem:** Switching between a chat app and a separate image tool slows down idea sketches made on the go. **The outcome:** An image generated from a text prompt in the chosen aspect ratio, with zoom and pan viewing and a download option. ### Steps 1. Install Pocket Strike on your Android device, following the build steps in the README. 2. Open a conversation in the app. 3. Type /image followed by your prompt and an aspect ratio, such as `/image futuristic electric sports car speeding through tokyo 16:9`. 4. Use the zoom and pan viewer to inspect the result. 5. Tap the download control to save the image to your gallery. ### Prompt to paste into your agent ```text In the PocketStrike repo, use the /image command in a conversation with the prompt and aspect ratio such as 9:16. Expected output: one generated image in the requested aspect ratio. Verify the aspect ratio of the image matches the one you typed before you save it. ``` ### Caveat The README lists SD-Turbo-Image-Gen at about 1.8 GB RAM and does not give generation times, so test with one prompt before planning larger batches. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/abuzar-ansarii-pocketstrike/generate-images-from-chat-prompts/ ## How to connect MCP servers to a ReAct agent on your phone with Pocket Strike Repo: [AbuZar-Ansarii/PocketStrike](https://github.com/AbuZar-Ansarii/PocketStrike) | Difficulty: Advanced | Time: 2 hours | Basis: From the README Add a Model Context Protocol server and let the agent use its tools with confirmation gates. **Who it is for:** A researcher or power user who wants an agent on their phone to search the web or query a service through MCP tools. **The problem:** Phone agents are limited unless they can reach outside tools. You want to add those tools while keeping control over destructive actions. **The outcome:** An agent that calls an MCP tool step by step, with reasoning logs and a confirmation prompt before any irreversible file action. ### Steps 1. Install Pocket Strike and open Settings, then the MCP area. 2. Open the MCP marketplace and add a built-in server such as Brave Search or GitHub, using Streamable HTTP or SSE transport. 3. Store any required API key in the app. Keys are kept in hardware-backed platform storage through flutter_secure_storage. 4. Set the confirmation policy to Always Ask under Security settings for the first runs. 5. Start an agent task and review each step in the reasoning log before approving file changes. ### Prompt to paste into your agent ```text In the PocketStrike repo, add the Brave Search MCP server from the marketplace and ask the ReAct agent to find three recent sources on . Expected output: a list of sources with the agent steps shown in the log. Verify that each tool call appears in the log before the final answer is given. ``` ### Caveat The README does not list the exact confirmation policy names in settings beyond Destructive Only, Always Ask, and Autonomous, so check the app screens before relying on a specific label. Risk, credentials: MCP servers such as GitHub may need API keys, so only add keys for services you trust and rotate them if the phone is lost. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. 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/abuzar-ansarii-pocketstrike/connect-mcp-servers-to-phone-agent/