summaryrefslogtreecommitdiffstats
path: root/README.md
diff options
context:
space:
mode:
authorsigoden <sigoden@gmail.com>2023-03-04 19:24:57 +0800
committersigoden <sigoden@gmail.com>2023-03-04 19:24:57 +0800
commit6dee2355970ded2e1de237fd2996d25ff01080db (patch)
tree9032477fa802a49ac34137363e5990b2340edf33 /README.md
parente594d1bd5f99d5a88135241d651995ea7a149394 (diff)
downloadaichat-6dee2355970ded2e1de237fd2996d25ff01080db.tar.gz
chore: release v0.2.0
Diffstat (limited to 'README.md')
-rw-r--r--README.md32
1 files changed, 16 insertions, 16 deletions
diff --git a/README.md b/README.md
index 2e1cf01..5eb54f3 100644
--- a/README.md
+++ b/README.md
@@ -5,7 +5,7 @@
Chat with ChatGPT-3.5 in the terminal.
-![demo](https://user-images.githubusercontent.com/4012553/222600858-3fb60051-2bf2-4505-92ff-649356cdb1f6.gif)
+![demo](https://user-images.githubusercontent.com/4012553/222897111-dd5015a0-abc1-4c65-a0fb-d491aba3c427.gif)
## Install
@@ -22,8 +22,8 @@ Download from [Github Releases](https://github.com/sigoden/aichat/releases), unz
## Features
- Compared to the browser, the terminal starts faster and needs less resources.
-- Support directive and interactive modes
-- Support markdown highlighting.
+- Interactive chat and imperative query.
+- Support highlight.
- Predefine prompts for role playing.
- History query/completion.
- Persist chat messages.
@@ -69,30 +69,30 @@ aichat may generate the following files in the config dir:
We can let ChatGPT play a certain role through `prompt` to make it better generate what we want. See [awesome-chatgpt-prompts](https://github.com/f/awesome-chatgpt-prompts) for details.
-We can predefine a batch of roles in `roles.yaml`. For example, we define a javascript-console role as follows.
+We can predefine a batch of roles in `roles.yaml`. For example, we define a emoji translator as follows.
```yaml
-- name: javascript-console
- prompt: >
- I want you to act as a javascript console. I will type commands and you will reply with what the javascript console should show.
- I want you to only reply with the terminal output inside one unique code block, and nothing else.
- do not write explanations. do not type commands unless I instruct you to do so.
- when i need to tell you something in english, i will do so by putting text inside curly brackets {like this}.
- My first command is:
+- name: emoji
+ prompt: >
+ I want you to translate the sentences I wrote into emojis. I will write the sentence, and you will express it with emojis.
+ I just want you to express it with emojis. I don't want you to reply with anything but emoji.
+ When I need to tell you something in English, I will do it by wrapping it in curly brackets like {like this}.
+ My first sentence is:
```
-Let ChaGPT answer questions in the role of a javascript-console.
+Let ChaGPT answer questions in the role of a emoji translator
```
-aichat --role javascript-console console.log("Hello World")
+$ aichat --role emoji I am very angry
+😠💢👿
```
-In interactive mode, we do this:
+In interactive chat, we do this:
```
-〉.role javascript-console
+〉.role emoji
-〉console.log("Hello world")
+〉I am very angry
```
## License