diff --git a/README.md b/README.md
index 2eb1c36..5b37de1 100644
--- a/README.md
+++ b/README.md
@@ -76,17 +76,19 @@ Options:
('print', 'click.echo', 'logging.info').
Default is click.echo.
--sub-model TEXT The sub-model name. Default is gpt-4o-mini.
+ --seed TEXT The random seed.
--log-level TEXT The log level, DEBUG, INFO, WARNING, ERROR
or CRITICAL. Default is WARNING.
--debug Whether to show debug logs or not.
--help Show this message and exit.
+
```
If you don't use `docker`, ignore `docker compose run werewolf` in the above commands.
## Examples
-### Case 1: 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day
+### Case1: 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day
This is an example of a werewolf game with 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day.
@@ -95,7 +97,7 @@ This is an example of a werewolf game with 1 villager, 1 werewolf, 1 knight and
Game Log
```sh
- $ docker compose run werewolf python -m werewolf -n 4 -w 1 -k 1 -f 1 -t 2 -m 'gpt-4o-mini' --sub-model 'gpt-4o-mini'
+$ docker compose run werewolf python -m werewolf -n 4 -w 1 -k 1 -f 1 -t 2 -m 'gpt-4o-mini' --sub-model 'gpt-4o-mini'
=============================== Day 1 (Daytime) ================================
Player0 (to GameMaster):
@@ -270,6 +272,7 @@ Think about what has happened and what you should do in the next action.
=============================================================================
================================ Game Result ================================
('VillagersWin', {'Player0': {'role': 'Knight', 'status': }, 'Player1': {'role': 'Werewolf', 'status': }, 'Player2': {'role': 'FortuneTeller', 'status': }, 'Player3': {'role': 'Villager', 'status': }})
+
```
@@ -456,7 +459,7 @@ Player1はゲームから除外されました。
-### Case 2: (Open Game) 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day
+### Case2: (Open Game) 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day
This is an example of a **opened** werewolf game with 1 villager, 1 werewolf, 1 knight and 1 fortune teller with 2 times to speak per day.
@@ -465,10 +468,10 @@ Note that each player cannot know the secret contents of what another player has
-Game Log
+ Game Log
```sh
- $ docker compose run werewolf python -m werewolf -n 4 -w 1 -k 1 -f 1 -t 2 -m 'gpt-4o-mini' --sub-model 'gpt-4o-mini' -o
+$ docker compose run werewolf python -m werewolf -n 4 -w 1 -k 1 -f 1 -t 2 -m 'gpt-4o-mini' --sub-model 'gpt-4o-mini' -o
GameMaster (to Player0):
You are a player with a role "ERole.Villager" in a werewolf game.
@@ -1398,6 +1401,7 @@ Think about what has happened and what you should do in the next action.
=============================================================================
================================ Game Result ================================
('WerewolvesWin', {'Player0': {'role': 'Villager', 'status': }, 'Player1': {'role': 'FortuneTeller', 'status': }, 'Player2': {'role': 'Werewolf', 'status': }, 'Player3': {'role': 'Knight', 'status': }})
+
```
@@ -2337,6 +2341,7 @@ Player0がゲームから除外されました。
+
## TODO
- [x] fix a bug that a night actions of a knight and a fortune teller are invalid. See [Case2](#case-2-open-game-1-villager-1-werewolf-1-knight-and-1-fortune-teller-with-2-times-to-speak-per-day);
@@ -2401,6 +2406,96 @@ Player0がゲームから除外されました。
Note that you can use `uv` to develop the project in step 3, 4 and 7.
+### Update README
+
+1. Fork the repository: [https://github.com/hmasdev/autogen-werewolf](https://github.com/hmasdev/autogen-werewolf)
+
+2. Clone the repository
+
+ ```bash
+ git clone https://github.com/{YOURE_NAME}/autogen-werewolf
+ cd autogen-werewolf
+ ```
+
+3. Create a virtual environment
+
+ ```bash
+ python -m venv venv
+ source venv/bin/activate
+ ```
+
+4. Install the required packages
+
+ ```bash
+ pip install -e .[dev]
+ ```
+
+5. Checkout your working branch
+
+ ```bash
+ git checkout -b your-working-branch
+ ```
+
+6. Make your changes
+
+ - Update **./README.md.j2**
+
+ - **CAUTION**: Do not update **./README.md** directly. It will be generated automatically from ./README.md.j2.
+
+ - Update or add an example in [./examples](./examples/)
+ - Name Rule: `Case{number}.json`
+ - Schema:
+
+ ```json
+ {
+ "$schema": "http://json-schema.org/draft-07/schema#",
+ "type": "object",
+ "properties": {
+ "title": {
+ "type": "string",
+ "description": "The title of the example case."
+ },
+ "description": {
+ "type": "string",
+ "description": "A brief description of the example case."
+ },
+ "log": {
+ "type": "string",
+ "description": "The game log for the example case."
+ },
+ "log_translated": {
+ "type": "object",
+ "description": "Translations of the game log.",
+ "additionalProperties": {
+ "type": "string"
+ }
+ }
+ },
+ "required": ["title", "description", "log", "log_translated"]
+ }
+ ```
+
+7. Test your changes
+
+ ```bash
+ python update_readme.py
+ ```
+
+8. Commit your changes
+
+ ```bash
+ git add .
+ git commit -m "Your commit message"
+ ```
+
+9. Push your changes
+
+ ```bash
+ git push origin your-working-branch
+ ```
+
+10. Create a pull request: [https://github.com/hmasdev/autogen-werewolf/compare](https://github.com/hmasdev/autogen-werewolf/compare)
+
### Classes
```mermaid