# AI use cases for Fde Gym (deeplethe/fde-gym) Repo: [deeplethe/fde-gym](https://github.com/deeplethe/fde-gym) | Use cases: 4 | Page: https://dedupedai.com/use-cases/deeplethe-fde-gym/ FDE Gym is an open online judge where each case is a simulated customer delivery that you work with a coding agent and are scored on once your system is live. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Practise customer delivery cases on your own machine: https://dedupedai.com/use-cases/deeplethe-fde-gym/practise-customer-delivery-cases-locally/ - Run a private FDE Gym site for your team's engineers: https://dedupedai.com/use-cases/deeplethe-fde-gym/run-a-private-training-site-for-a-team/ - Work an FDE Gym case from the command line with the harness: https://dedupedai.com/use-cases/deeplethe-fde-gym/work-a-case-from-the-command-line/ - Add extra runner machines to a self-hosted FDE Gym site: https://dedupedai.com/use-cases/deeplethe-fde-gym/add-runner-machines-to-a-self-hosted-site/ ## How to practise customer delivery cases on your own machine with Fde Gym Repo: [deeplethe/fde-gym](https://github.com/deeplethe/fde-gym) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Run FDE Gym locally and work the three open cases with a coding agent. **Who it is for:** An engineer preparing for forward deployed engineer work who wants realistic customer delivery practice before applying or starting a role. **The problem:** Most coding practice gives a complete, exact spec. Real customer work starts with a vague request, a running system and people to ask. **The outcome:** A local FDE Gym site with the three open cases loaded, where you can open a case, read what the customer handed over, and start a run. ### Steps 1. Install Docker and Python 3.9 or later, then create an OpenRouter key and keep it in a file outside the repository. 2. `git clone https://github.com/deeplethe/fde-gym.git && cd fde-gym` 3. `mkdir -p ~/.fdegym && echo "sk-or-..." > ~/.fdegym/openrouter_key` 4. `python3 scripts/fetch_cases.py` 5. `docker compose up --build` 6. Open http://127.0.0.1:8787, sign up first so your account becomes the admin, pick a case and press Begin. ### Prompt to paste into your agent ```text In the fde-gym repository at , follow the README quick start to download the three open cases into cases/ and start the site with docker compose. Report the URL the site answers on and the names of the case folders present. Verify that cases/ contains the three downloaded cases before reporting success. ``` ### Caveat The site listens only on your own machine by default, and the README warns to read the Security section before letting anyone else in. Risk, credentials: The OpenRouter key is read from a file in your home directory, so keep that file out of version control and out of shared logs. 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/deeplethe-fde-gym/practise-customer-delivery-cases-locally/ ## How to run a private FDE Gym site for your team's engineers with Fde Gym Repo: [deeplethe/fde-gym](https://github.com/deeplethe/fde-gym) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README Host the judge for your own people, with an admin console for cases and members. **Who it is for:** A engineering lead or training owner at a company who wants new delivery engineers to practise customer work on cases the team controls. **The problem:** Public practice sites do not show your own delivery patterns, and you need control over who signs up and which cases are visible. **The outcome:** A team-run FDE Gym instance with an admin account, uploaded cases, and a leaderboard and submissions view for your members. ### Steps 1. `git clone https://github.com/deeplethe/fde-gym.git && cd fde-gym` 2. `docker compose up --build` 3. Sign up first on the site so your account becomes the admin, then open the admin console. 4. Upload or file a case in the admin console, and pause or remove any case that is not ready. 5. Read docs/self-hosting.md for settings, managed database and object store options, and the Security section before you invite anyone. ### Prompt to paste into your agent ```text In the fde-gym repository at , read docs/self-hosting.md and list the settings a team needs to change before sharing the site. Output the list as bullet points with the setting name and what it controls. Verify each item appears in that document before listing it. ``` ### Caveat The README says the site executes the code people write, so a shared deployment needs the isolation steps in docs/self-hosting.md. Risk, credentials: Admin accounts control members, runs and the model spending cap, so choose the first sign-up with care and protect the admin password. 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/deeplethe-fde-gym/run-a-private-training-site-for-a-team/ ## How to work an FDE Gym case from the command line with the harness with Fde Gym Repo: [deeplethe/fde-gym](https://github.com/deeplethe/fde-gym) | Difficulty: Advanced | Time: 30 minutes | Basis: From the README Use the Python harness to work a case without the browser workbench. **Who it is for:** A developer who prefers terminal workflows, or who wants to hand a case to an AI agent that cannot use the browser. **The problem:** The browser workbench is the main route into a case, but an agent or a terminal user needs the same case from a script or shell. **The outcome:** A confirmed harness install and a clear path to running one case from the command line, with the workflow documented in harness/README.md. ### Steps 1. `git clone https://github.com/deeplethe/fde-gym.git && cd fde-gym` 2. `python3 harness/run.py --version` 3. Open harness/README.md and read how to work a case from the command line. 4. Follow the case-running instructions in harness/README.md against one case folder from cases/. (inferred, check the README) ### Prompt to paste into your agent ```text In the fde-gym repository at , run `python3 harness/run.py --version` and read harness/README.md. Report the harness version and the exact command the README gives for working a case. Verify the version string comes from the command output, not from the README. ``` ### Caveat The README only names the harness as Python standard library code and points to harness/README.md for how to run a case, so check that file before relying on any command. 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/deeplethe-fde-gym/work-a-case-from-the-command-line/ ## How to add extra runner machines to a self-hosted FDE Gym site with Fde Gym Repo: [deeplethe/fde-gym](https://github.com/deeplethe/fde-gym) | Difficulty: Advanced | Time: 60 minutes | Basis: From the README Attach more machines to run cases, on any network that reaches the site. **Who it is for:** An operator of a team FDE Gym site whose runs are queuing or whose cases need more capacity than one machine provides. **The problem:** A single runner can become a bottleneck when many people work cases at once, and the site needs more machines without sharing its database or keys. **The outcome:** Additional runners attached to the site, with a clear record of which machine runs which case work. ### Steps 1. Read docs/self-hosting.md and its section on adding machines to run cases on. 2. On a second machine, follow the runner setup in docs/self-hosting.md so it can reach the site over the network. (inferred, check the README) 3. Confirm the new runner appears in the admin console's machines list. 4. Start a test run and check that its terminal and grading work on the new machine. ### Prompt to paste into your agent ```text In the fde-gym repository at , read the runner section of docs/self-hosting.md and list the network and setup requirements for an extra runner. Output a numbered checklist. Verify each item is stated in that document before adding it. ``` ### Caveat The README says runners hold no database, storage or model key, but a runner still executes case workloads, so place it on a machine you are willing to run this code on. 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/deeplethe-fde-gym/add-runner-machines-to-a-self-hosted-site/