# AI use cases for Vibe Service Guardian (keith-yan/vibe-service-guardian) Repo: [keith-yan/vibe-service-guardian](https://github.com/keith-yan/vibe-service-guardian) | Use cases: 3 | Page: https://dedupedai.com/use-cases/keith-yan-vibe-service-guardian/ Vibe Service Guardian is a local-first Python tool that links listening services to their project and agent, flags stale dev servers, monitors local AI runtimes, and estimates which open-weight models a machine can run. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Find which project owns a stale dev server and stop it safely: https://dedupedai.com/use-cases/keith-yan-vibe-service-guardian/find-and-retire-stale-dev-servers/ - Check whether an open-weight model fits your machine before you download it: https://dedupedai.com/use-cases/keith-yan-vibe-service-guardian/check-if-a-model-fits-before-downloading/ - Run the guardian as a portable Windows app from a zip file: https://dedupedai.com/use-cases/keith-yan-vibe-service-guardian/run-the-guardian-as-a-portable-windows-app/ ## How to find which project owns a stale dev server and stop it safely with Vibe Service Guardian Repo: [keith-yan/vibe-service-guardian](https://github.com/keith-yan/vibe-service-guardian) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Trace a listening port to its project and agent, then stop only ordinary host dev processes after confirmation. **Who it is for:** A developer whose laptop has several forgotten dev servers, agent-started processes, or ports held by old runs, and who needs to know which ones are safe to stop. **The problem:** Listening ports do not say which project, agent, or terminal started them, or whether they were left behind. Killing the wrong process can break other work. **The outcome:** You see each service with its project, agent or terminal evidence, and a stale or not-stale call backed by visible evidence. Confirmed stale processes are stopped, and the port and PID are checked afterward. ### Steps 1. Create a virtual environment with `python -m venv .venv` and install the runtime from `requirements-lock` as described in the README source-run section. 2. Start the console with `python -m vsg --open` and open the services monitor. 3. Open a service detail and compare its project, agent or terminal attribution, and the listed evidence. 4. Record whether the service is stale, not stale, or unsure, so the decision is kept in the local operation log. 5. For a confirmed stale ordinary host dev process, type `STOP ` in the confirmation field, then check that the original port and PID are gone. ### Prompt to paste into your agent ```text Using the repo keith-yan/vibe-service-guardian, run `python -m vsg --open` and inspect the service on port . Report its PID, project attribution, and the stale-check evidence. Verify that you do not stop anything unless the service is an ordinary host dev process and the user has typed the stop confirmation. ``` ### Caveat Stop is only enabled for ordinary host dev or inference processes; Agent processes, Windows services, Docker, WSL, and protected targets stay read-only. 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/keith-yan-vibe-service-guardian/find-and-retire-stale-dev-servers/ ## How to check whether an open-weight model fits your machine before you download it with Vibe Service Guardian Repo: [keith-yan/vibe-service-guardian](https://github.com/keith-yan/vibe-service-guardian) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Compare physical load limits, practical limits, and target SLA limits for a planned user count and concurrency. **Who it is for:** A developer or team lead deciding whether a local open-weight model can serve a planned number of users on a specific laptop, workstation, or GPU box. **The problem:** Model file size alone does not show whether KV cache, concurrency, token speed, and time to first token will fit in the memory that is actually free. **The outcome:** You get three separate answers: the physical load ceiling, the practical ceiling with a minimum speed, and the ceiling that meets your target. Each candidate shows weights, KV cache, workspace, and risk reasons. ### Steps 1. Start the console with `python -m vsg --open`. 2. Open the model capacity page and let it read CPU, memory, disk, GPU or unified memory, and installed local runtimes. 3. Enter planned users, peak concurrency, average input and output tokens, per-session context, target tokens per second per user, and target time to first token. 4. Review each candidate's weights, KV cache, workspace, per-user throughput range, and the maximum concurrency under target. 5. Copy the generated run command template, which is fixed to 127.0.0.1 and is not executed automatically, and test it on the real machine before relying on it. ### Prompt to paste into your agent ```text Using keith-yan/vibe-service-guardian, open the model capacity page with `python -m vsg --open`, enter the plan for users at concurrency, and report the physical, practical, and target SLA ceilings for . Verify that each figure lists its risk reasons and that the result is treated as an estimate until a local benchmark is run. ``` ### Caveat The capacity output is a planning estimate and does not replace a same-load benchmark, and the built-in model catalog is a dated offline snapshot rather than a full list. 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/keith-yan-vibe-service-guardian/check-if-a-model-fits-before-downloading/ ## How to run the guardian as a portable Windows app from a zip file with Vibe Service Guardian Repo: [keith-yan/vibe-service-guardian](https://github.com/keith-yan/vibe-service-guardian) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Build a portable Windows x64 zip from the source folder and start or stop the console with included launchers. **Who it is for:** A Windows user who wants a self-contained copy of the tool they can keep in one folder without installing Python for daily use. **The problem:** Setting up Python and a virtual environment every time is a barrier, and a shared machine may not allow installs. **The outcome:** You get a portable zip with Start, Open, and Stop launchers, and runtime data stays in the portable folder's data directory. ### Steps 1. In the source folder, run `Build-Portable.cmd` to generate `Vibe-Service-Guardian-Windows-x64-0.8.5.2.zip`. 2. Extract the zip to a writable folder that only the current user can access. 3. Double-click `Start-VSG.cmd` to start the console. 4. If the browser does not open on its own, double-click `Open-VSG.cmd`. 5. When you are done, double-click `Stop-VSG.cmd`, and check the `.sha256` file that ships with the package. ### Prompt to paste into your agent ```text Using the repo keith-yte/vibe-service-guardian, run `Build-Portable.cmd` in the source folder at , then confirm the zip exists with a matching `.sha256` file. Verify that the console starts through `Start-VSG.cmd` and stops through `Stop-VSG.cmd`. ``` ### Caveat The EXE is unsigned, so Windows SmartScreen may show an unknown publisher warning, and the portable build is not a signed release. 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/keith-yan-vibe-service-guardian/run-the-guardian-as-a-portable-windows-app/