Demo: four deciders to learn the question types

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

# Demo: four deciders to learn the question types

*[Version française](README.fr.md)*

The four examples of [How to write the questions of a decider](../../docs/en/how-to/write-decision-questions.md), as agent files you can run. Each one is a **decider** (LiquidAI d1-3B on llmman, `POST /v1/systemone`): no chat model, it answers typed questions about the message it receives.

| File | Decides | Questions |
|------|---------|-----------|
| `symptoms.yaml` | how serious a symptom ticket is | `severity` (score), `red_flag` (noul), `who` (choice) |
| `dice421.yaml` | how many dice to throw again at 421 | `reroll` (choice) |
| `spell.yaml` | which spell a wizard should cast | `spell` (choice) |
| `fight.yaml` | fight, flee or talk | `action` (choice), `danger` (score), `hurt` (noul) |

## Run it

Pull the model into llmman once:

```bash
llmman pull hf.co/LiquidAI/d1-3B-GGUF:Q4_K_M
```

Then, from this directory:

```bash
mm -config spell.yaml -p "An ogre is raising its club to hit me right now."   # the server's JSON
mm -config fight.yaml                                                         # type situations, read one line per question
```

More messages to try, and the answers measured on 2026-10-24, are in [the guide](../../docs/en/how-to/write-decision-questions.md#examples). For `dice421.yaml`, write what is missing into the message (`Dice: 4, 2, 6. In place for 421: 4 and 2. Missing: 1.`): given the dice alone, the decider got it wrong.