# AI use cases for Gear (rsi-gear/gear)

Repo: [rsi-gear/gear](https://github.com/rsi-gear/gear) | Use cases: 3 | Page: https://dedupedai.com/use-cases/rsi-gear-gear/

Gear is an open-source framework that iteratively improves an AI agent's harness (instructions, tools, skills, and workflows) using evaluation results from a benchmark of tasks.

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

- Tune a marketing agent on AutomationBench tasks: https://dedupedai.com/use-cases/rsi-gear-gear/tune-marketing-agent-automationbench/
- Optimize an agent against your own task set: https://dedupedai.com/use-cases/rsi-gear-gear/optimize-agent-on-own-task-set/
- Set a weighted objective that balances pass rate and cost: https://dedupedai.com/use-cases/rsi-gear-gear/balance-pass-rate-against-cost/

## How to tune a marketing agent on AutomationBench tasks with Gear

Repo: [rsi-gear/gear](https://github.com/rsi-gear/gear) | Difficulty: Intermediate | Time: 2 hours | Basis: From the README

Run one Gear optimization round on AutomationBench Marketing tasks to improve an agent's pass rate.

**Who it is for:** A marketing operations lead who wants an agent to handle Marketing automation tasks more reliably and has no benchmark of their own yet.

**The problem:** A general agent setup completes only part of the Marketing tasks, and tuning instructions and tools by hand is slow and hard to measure.

**The outcome:** A proposed round of harness changes, evaluated on the Marketing tasks, with each harness version and its results recorded so you can inspect and reuse the best one.

### Steps

1. Run `npm install --global rsi-gear@latest agent-hitch@latest @deepseek-ai/dsh@latest` to install Gear, the Hitch evaluation runner, and DSH.
2. Run `hitch eval setup harbor` to set up the Harbor evaluation runner.
3. Install the bundled Refine Skill in your agent environment by following the quick start guide, then configure AutomationBench as the benchmark.
4. In your agent session, invoke `/refine` or submit a request to improve performance on AutomationBench's Marketing tasks and run one round of optimization.
5. Review the recorded harness versions and their evaluation results, then keep the version that scores best for your use.

### Prompt to paste into your agent

```text
Use the Gear repo at rsi-gear/gear to run one Refine Skill optimization round on AutomationBench's Marketing tasks for <target agent setup>. Report the pass rate before and after the round and the path to the recorded harness versions. Verify that each retained harness version has a matching evaluation result before you report success.
```

### Caveat

Gear optimizes the harness only; the model training loop is experimental, so this use case changes instructions, tools, and workflows and does not retrain the model.

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/rsi-gear-gear/tune-marketing-agent-automationbench/

## How to optimize an agent against your own task set with Gear

Repo: [rsi-gear/gear](https://github.com/rsi-gear/gear) | Difficulty: Advanced | Time: 3 hours | Basis: From the README

Package your own tasks in Harbor format and run Gear's Refine Skill to improve the agent on them.

**Who it is for:** A developer who has a repeatable set of agent tasks with clear pass or fail criteria and wants the agent to get better at that specific work.

**The problem:** No public benchmark matches your workflow, so the agent's failures on your real tasks go unmeasured and unfixed.

**The outcome:** A harness tuned to your task set, with each version and its results recorded so you can compare changes and reuse the best harness.

### Steps

1. Build your task set in Harbor format by following the dataset guide, with a clear success criterion for each task.
2. Run `npm install --global rsi-gear@latest agent-hitch@latest @deepseek-ai/dsh@latest` to install Gear and its runners.
3. Run `hitch eval setup harbor` to set up the Harbor evaluation runner.
4. Install the Refine Skill in your agent environment and point its configuration at your task paths, target harness, and model settings.
5. Invoke `/refine` where supported and run one optimization round, then review the recorded harness versions and their results.

### Prompt to paste into your agent

```text
Use the Gear repo at rsi-gear/gear to optimize the harness of <target agent> on my task set at <task path in Harbor format>. Report the pass rate for the starting harness and for the best retained harness. Verify that the task paths resolve and that the Refine Skill connects to Gear before the first run.
```

### Caveat

Results are only as good as the tasks, so the success criteria in your Harbor tasks must be clear enough for Gear to score 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/rsi-gear-gear/optimize-agent-on-own-task-set/

## How to set a weighted objective that balances pass rate and cost with Gear

Repo: [rsi-gear/gear](https://github.com/rsi-gear/gear) | Difficulty: Intermediate | Time: 2 hours | Basis: From the README

Combine pass rate, process scores, and cost or tokens into one objective when tuning a harness.

**Who it is for:** A team lead who must keep agent quality high while limiting API spend and wants to pick the harness that gives the best trade-off.

**The problem:** Optimizing only for pass rate can produce a harness that is correct but expensive to run on every task.

**The outcome:** A harness selected by a weighted objective, with the original pass rate, process scores, and cost measurements still visible for each version.

### Steps

1. Read the objective guide to see how declared pass rate, process scores, cost, and tokens can be weighted.
2. Define your weights for pass rate, process scores, and cost or tokens for your task set.
3. Run one Refine Skill optimization round on your task set with that weighted objective.
4. Compare the retained harness with the starting harness on pass rate and on cost or tokens, using the recorded results.

### Prompt to paste into your agent

```text
Use the Gear repo at rsi-gear/gear to define a weighted objective that combines pass rate and cost for <target task set> and run one optimization round with it. Report the weights used and the pass rate and cost of the retained harness against the starting harness. Verify that the original measurements are still present in the results.
```

### Caveat

The objective guide defines how weights combine measurements, so choose weights that match your own priorities rather than copying the Marketing experiment's settings.

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/rsi-gear-gear/balance-pass-rate-against-cost/