# AI use cases for DSH Agent Team Gui (toolclub/dsh-agent-team-gui)

Repo: [toolclub/dsh-agent-team-gui](https://github.com/toolclub/dsh-agent-team-gui) | Use cases: 4 | Page: https://dedupedai.com/use-cases/toolclub-dsh-agent-team-gui/

A DeepSeek Harness Web profile plugin that saves reusable multi-model agent teams, runs planned workflows with per-member models and tools, and reports run status and Token usage.

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

- Run a planner, engineer, and reviewer team on a small task: https://dedupedai.com/use-cases/toolclub-dsh-agent-team-gui/run-full-stack-delivery-team/
- Save a team once and reuse it across projects: https://dedupedai.com/use-cases/toolclub-dsh-agent-team-gui/reuse-team-across-projects/
- Check how many Tokens a multi-agent run used: https://dedupedai.com/use-cases/toolclub-dsh-agent-team-gui/audit-team-token-usage/
- Run a long task in the background with a review gate: https://dedupedai.com/use-cases/toolclub-dsh-agent-team-gui/run-long-task-in-background-with-review/

## How to run a planner, engineer, and reviewer team on a small task with DSH Agent Team Gui

Repo: [toolclub/dsh-agent-team-gui](https://github.com/toolclub/dsh-agent-team-gui) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README

Import the example recipe and have three team members build and test a small app in a scratch project.

**Who it is for:** A developer who wants to see a planner, implementer, and reviewer work together on a real coding task before building their own team.

**The problem:** A single chat agent can write code but gives no visible plan, separate review, or record of each step's output and cost.

**The outcome:** A tested small project, plus a Run Center record of the plan, each member's output, the review, and actual Token coverage.

### Steps

1. Download examples/full-stack-delivery.recipe.json and open Settings → Teams → Recipes & data, then choose Choose recipe file.
2. Map the placeholder routes to your configured provider/model routes, review the preview, and choose Create a copy → Import reviewed recipe.
3. Set Team usage to Host dispatches first, Dispatch policy to Run assigned work, and member selection to All members, then save.
4. Select an empty scratch project, choose the imported team beside the composer, select Select team for this conversation, and send the ready-to-copy to-do list task from docs/first-team.md.
5. Open Team runs to inspect the plan, outputs, review, and Token coverage, then check the delivered files and test results yourself.

### Prompt to paste into your agent

```text
Using the dsh-agent-team-gui repo, import examples/full-stack-delivery.recipe.json into an empty scratch project at <project-path> and run the to-do list task from docs/first-team.md. Report the Team runs summary with the plan, member outcomes, and Token coverage. Verify that the built project's test command passes before reporting success.
```

### Caveat

The recipe uses placeholder routes that must be mapped to your own DSH provider/model routes before the first run.

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/toolclub-dsh-agent-team-gui/run-full-stack-delivery-team/

## How to save a team once and reuse it across projects with DSH Agent Team Gui

Repo: [toolclub/dsh-agent-team-gui](https://github.com/toolclub/dsh-agent-team-gui) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README

Define a planner, implementer, and reviewer with their own models, then select the team per conversation or per project.

**Who it is for:** A developer who repeats the same multi-step workflow across several repos and wants one saved team instead of a new dispatch each time.

**The problem:** Re-creating the same agent roles and model choices for every project costs time and produces inconsistent results.

**The outcome:** A named team in Settings → Teams that you can attach to any project or conversation with the Team, Solo, or Inherited modes.

### Steps

1. In Settings → Teams → Member library, save each member's role, model, optional fallback, and tool policy.
2. In Settings → Teams, add those members and choose activation, member selection, and execution options.
3. Leave Fixed order off for dynamic planning, or enable it if you need a fixed serial sequence.
4. Choose Team, Solo, or Inherited beside the composer to control the current conversation.
5. Run a short test task in one project, then reuse the same team in a second project to confirm the selection holds.

### Prompt to paste into your agent

```text
Using the dsh-agent-team-gui repo, create a team named <team-name> in Settings → Teams with a planner, an implementer, and a reviewer, each with its own model route. Report the saved member list and the activation settings. Verify that the team still appears after a page refresh.
```

### Caveat

Team members use the model routes you assign in DSH, so a team only works in projects where those routes are configured.

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/toolclub-dsh-agent-team-gui/reuse-team-across-projects/

## How to check how many Tokens a multi-agent run used with DSH Agent Team Gui

Repo: [toolclub/dsh-agent-team-gui](https://github.com/toolclub/dsh-agent-team-gui) | Difficulty: Beginner | Time: 30 minutes | Basis: From the README

Read per-run Token buckets and coverage status in Run Center for planner, member, review, and repair usage.

**Who it is for:** A developer or team lead who needs to know how much model usage a multi-agent task consumed before running similar work again.

**The problem:** Multi-agent runs spread usage across planning, members, and review, which makes it hard to tell where Tokens went.

**The outcome:** A per-run and per-attempt breakdown of uncached input, cache read, cache write, and output Tokens, with coverage marked full, partial, or unavailable.

### Steps

1. Run any team task, or open an existing run in Team runs.
2. Open the run's usage view and read the four buckets: uncached input, cache read, cache write, and output.
3. Check the coverage label for the run and for each retry attempt, and note any partial or unavailable entries.
4. Compare planner, member, review, and repair usage to see which step used the most Tokens.
5. Do not convert Tokens into money, because the plugin does not guess prices.

### Prompt to paste into your agent

```text
Using the dsh-agent-team-gui repo, open the Team runs entry for <run-id> and extract the planner, member, review, and repair Token counts with their coverage status. Report a table of the four buckets per step. Verify that any run showing Metering… is reported as pending, not as zero.
```

### Caveat

Token figures come from provider-reported usage, so runs with partial or unavailable coverage will not show a complete total.

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/toolclub-dsh-agent-team-gui/audit-team-token-usage/

## How to run a long task in the background with a review gate with DSH Agent Team Gui

Repo: [toolclub/dsh-agent-team-gui](https://github.com/toolclub/dsh-agent-team-gui) | Difficulty: Advanced | Time: 90 minutes | Basis: From the README

Start a long team task that keeps running in the Run Center, with a reviewer and up to two repair rounds.

**Who it is for:** A developer starting a large refactor or documentation pass who does not want to keep the chat open while it runs.

**The problem:** Long agent work blocks the conversation, and a rejected output can be lost if nothing checks it.

**The outcome:** A background run that stays visible in Run Center, passes through a named reviewer, and can be cancelled or retried if it stops.

### Steps

1. In the team settings, enable the optional quality gate and name one reviewer, one repair owner, and explicit criteria, with 0..2 repair rounds.
2. Send the long task with the team selected and choose background execution.
3. Watch the live phase, elapsed time, and child IDs in Run Center.
4. When a rejection occurs, read the reviewer's reason and let the named repair owner rerun.
5. If a host restart interrupts the run, check that Run Center marks it Interrupted rather than Completed, then start a linked retry.

### Prompt to paste into your agent

```text
Using the dsh-agent-team-gui repo, start a background team run for <task-description> in <project-path> with a quality gate of 1 repair round. Report the Run Center status, final quality gate decision, and any Interrupted runs. Verify that the run's final state matches the reviewer's accepted output.
```

### Caveat

Background runs only survive a Host restart as Interrupted when the optional DSH Jobs service is absent, so the cancel hook and resume behavior depend on your DSH setup.

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/toolclub-dsh-agent-team-gui/run-long-task-in-background-with-review/