# AI use cases for Fuxi (fuxicodex/Fuxi)

Repo: [fuxicodex/Fuxi](https://github.com/fuxicodex/Fuxi) | Use cases: 4 | Page: https://dedupedai.com/use-cases/fuxicodex-fuxi/

FuXi is a fast terminal AI coding agent that reads code, edits files, runs commands, and drives tools, with cost-aware routing across LLM providers.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Fix failing tests in a repo from the terminal: https://dedupedai.com/use-cases/fuxicodex-fuxi/fix-failing-tests-from-terminal/
- Run FuXi on your own provider key and switch models: https://dedupedai.com/use-cases/fuxicodex-fuxi/use-own-provider-key-and-switch-models/
- Set up FuXi on a headless machine with a setup token: https://dedupedai.com/use-cases/fuxicodex-fuxi/set-up-headless-machine-with-setup-token/
- Run FuXi with bypassed approvals inside a container: https://dedupedai.com/use-cases/fuxicodex-fuxi/run-agent-permissions-in-isolated-sandbox/

## How to fix failing tests in a repo from the terminal with Fuxi

Repo: [fuxicodex/Fuxi](https://github.com/fuxicodex/Fuxi) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Point FuXi at a repository with failing tests and have it read, edit, and rerun until they pass.

**Who it is for:** A developer whose project has failing tests and who wants a terminal agent to investigate and patch them.

**The problem:** Tracing a failing test through the code and making a correct fix takes time, especially in an unfamiliar repository.

**The outcome:** FuXi reads the relevant code, proposes or applies edits, runs commands, and reports the changes it made.

### Steps

1. `curl -fsSL https://downloads.fuxicode.com/bootstrap.sh | bash`
2. `fuxi login`
3. `fuxi doctor`
4. `cd <your-repo>` (inferred, check the README)
5. `fuxi`, then type: Fix the failing tests in this repository.

### Prompt to paste into your agent

```text
In the repository at <repo-path>, run FuXi from the terminal with the prompt 'Fix the failing tests in this repository.' Expected output: a summary of the files FuXi changed and the result of its test run. Verify by running the project's own test command yourself and confirming the previously failing tests now pass.
```

### Caveat

Sign-in gives access to FuXi-managed models, and the README says registration processes minimal account data, so check the privacy docs before using it on private code.

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/fuxicodex-fuxi/fix-failing-tests-from-terminal/

## How to run FuXi on your own provider key and switch models with Fuxi

Repo: [fuxicodex/Fuxi](https://github.com/fuxicodex/Fuxi) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Connect your own API key and change models with /model to manage which provider handles your work.

**Who it is for:** A developer who already pays for an LLM provider and wants FuXi to use that key instead of FuXi-managed models.

**The problem:** Teams often want to control which provider and model handle their code, and to change models without restarting their workflow.

**The outcome:** FuXi sends requests to the provider you chose and lets you switch models in the middle of a session.

### Steps

1. `fuxi init`
2. Set your API key in `~/.fuxi/config.yaml` or as an environment variable, as described in the usage guide.
3. `fuxi doctor`
4. `fuxi`
5. Type `/model` in the session to pick a different model.

### Prompt to paste into your agent

```text
Configure the FuXi repo at fuxicodex/Fuxi to use the API key for <provider-name> from the local environment, then start a session. Expected output: FuXi starts and answers a short test prompt using the provider you set. Verify that the request is billed to your own provider account and not to FuXi-managed models.
```

### Caveat

The README does not list which LLM providers are supported, so check the environment variable reference before assuming your provider works.

Risk, credentials: Your provider API key is stored in local config or an environment variable, so keep the config file out of shared repositories and 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/fuxicodex-fuxi/use-own-provider-key-and-switch-models/

## How to set up FuXi on a headless machine with a setup token with Fuxi

Repo: [fuxicodex/Fuxi](https://github.com/fuxicodex/Fuxi) | Difficulty: Advanced | Time: 30 minutes | Basis: From the README

Authenticate FuXi on a server or CI runner without an interactive sign-in using a setup token.

**Who it is for:** A platform engineer who needs FuXi available on a headless server or CI runner where no browser sign-in is possible.

**The problem:** Interactive login does not work on servers or CI machines, so the agent cannot be authenticated there.

**The outcome:** The headless machine is authenticated with a setup token and FuXi's environment checks pass.

### Steps

1. `curl -fsSL https://downloads.fuxicode.com/bootstrap.sh | bash`
2. `fuxi setup-token`
3. Store the issued token in your CI secret store (inferred, the README does not describe storage).
4. `fuxi doctor`
5. `fuxi --version`

### Prompt to paste into your agent

```text
On the headless host <host-name>, install FuXi from the fuxicodex/Fuxi repo and authenticate with `fuxi setup-token`. Expected output: `fuxi doctor` completes its environment checks without errors. Verify by running `fuxi --version` and confirming the version prints.
```

### Caveat

The README only says setup-token is for headless and CI use and does not document a non-interactive prompt mode, so verify how your pipeline should call FuXi.

Risk, credentials: A setup token grants access to your FuXi account, so store it as a secret and do not print it in build logs.

Risk, remote script: The install command pipes a remote bootstrap script directly into bash, so review the script or pin a known version before running it on a server.

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/fuxicodex-fuxi/set-up-headless-machine-with-setup-token/

## How to run FuXi with bypassed approvals inside a container with Fuxi

Repo: [fuxicodex/Fuxi](https://github.com/fuxicodex/Fuxi) | Difficulty: Advanced | Time: 30 minutes | Basis: From the README

Let FuXi run tool calls without prompts, but only inside a disposable container or VM.

**Who it is for:** A developer doing a large automated refactor who wants fewer approval prompts and has an isolated environment to run it in.

**The problem:** Approving every tool call slows long refactors, but unrestricted execution on a main workstation is risky.

**The outcome:** FuXi runs its tool calls without prompts inside an isolated environment, and the host machine stays protected.

### Steps

1. Start a disposable container or VM and open a shell inside it. (inferred, check the README)
2. `curl -fsSL https://downloads.fuxicode.com/bootstrap.sh | bash`
3. `fuxi --dangerously-skip-permissions` (inferred, check the README)
4. Type your refactor prompt and review the changes when the session ends.
5. Remove the container or VM when the work is done.

### Prompt to paste into your agent

```text
Inside a disposable container or VM, install FuXi from fuxicodex/Fuxi and start it with `fuxi --dangerously-skip-permissions` in the repository at <repo-path>. Expected output: the session starts in bypassPermissions mode and completes the requested change. Verify by checking the diff in the repository and confirming nothing outside the sandbox changed.
```

### Caveat

The README says root and sudo are rejected on Linux and macOS, and that this flag should be used only in a container or VM, so do not run it on your main workstation.

Risk, remote script: The install step pipes a remote bootstrap script into bash, so review it before running it inside your sandbox.

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/fuxicodex-fuxi/run-agent-permissions-in-isolated-sandbox/