chatsnack

Name: chatsnack
Author: Mattie

29stars2forks

How to install chatsnack?

mattie/chatsnack

$git clone https://github.com/Mattie/chatsnack

Installs into the current project.

Summary

chatsnack is the easiest Python library for rapid development with OpenAI's ChatGPT API. It provides an intuitive interface for creating and managing chat-based prompts and responses, making it convenient to build complex, interactive conversations with AI.

Files · 1

View on GitHub

README.md

1	# chatsnack
2
3	chatsnack is the easiest Python library for rapid development with OpenAI's ChatGPT API. It provides an intuitive interface for creating and managing chat-based prompts and responses, making it convenient to build complex, interactive conversations with AI.
4
5	[Documentation site](https://mattie.github.io/chatsnack/)
6
7	![chatsnack features](/docs/chatsnack_features_smaller.jpg)
8	## Setup
9
10	### Got snack?
11
12	Install the `chatsnack` package from PyPI:
13
14	```bash
15	# requires Python 3.11+
16	pip install chatsnack
17	```
18
19	### Got keys?
20
21	Add your OpenAI API key to your .env file. If you don't have a .env file, the library will create a new one for you in the local directory.
22
23	### Learn More!
24	Read more below, watch the [intro video](https://www.youtube.com/watch?v=Yjwi54rHrhw) or check out the [Getting Started notebook](notebooks/).
25
26	### Responses API Support
27	`Chat()` now defaults to the Responses family over WebSocket with `session="inherit"`.
28	If you stay on that default path--or explicitly choose another Responses transport--
29	chatsnack requires the OpenAI Python client to be `openai>=2.29.0` (or newer).
30	For explicit WebSocket support outside chatsnack's packaged dependencies, install
31	`openai[realtime]>=2.29.0`.
32
33	## Usage
34
35	### Enjoy a Quick Snack
36
37	Easiest way to get going with `chatsnack` is with built-in snack packs. Each pack is a singleton ready to mingleton.
38
39	```python
40	>>> from chatsnack.packs import ChatsnackHelp
41	>>> ChatsnackHelp.ask("What is your primary directive?")
42	```
43	> *"My primary directive is to assist users of the chatsnack Python module by answering questions
44	and helping with any problems or concerns related to the module. I aim to provide helpful and
45	informative responses based on the chatsnack module's documentation and best practices."*
46
47	You can try out other example snack packs like `Confectioner`, `Jolly`, `Chester`, `Jane`, or `Data`. (Eventually there will be an easy way to create and share your own.)
48
49	Instead of `.ask()` we can call `.chat()` which will allow us to continue a conversation.
50	```python
51	>>> mychat = ChatsnackHelp.chat("What is chatsnack?") # submits and returns a new chat object
52	>>> print(mychat.response)
53	```
54	> Chatsnack is a Python module that provides a simple and powerful interface for creating conversational agents and tools using OpenAI's ChatGPT language models. It allows you to easily build chat prompts, manage conversation flow, and integrate with ChatGPT to generate responses. With Chatsnack, you can create chatbots, AI-assisted tools, and other conversational applications using a convenient and flexible API.
55
56	Now we can add more messages to that chat however we'd like:
57	```python
58	>>> mychat.user("Respond in only six word sentences from now on.")
59	>>> mychat.asst("I promise I will do so.")
60	>>> mychat.user("How should I spend my day?")
61	>>> mychat.ask()
62	```
63	> "Explore hobbies, exercise, connect with friends."
64
65	If you want a super simple interactive conversation with you and a chatbot, you could do something like this:
66	```python
67	from chatsnack.packs import Jolly
68	yourchat = Jolly # interview a green giant
69	while (user_input := input("Chat with the bot: ")):
70	print(f"USER: {user_input}")
71	yourchat = yourchat.chat(user_input)
72	print(f"THEM: {yourchat.last}")
73	```
74
75	### Natural Attachments (Phase 3 style ergonomics)
76
77	For the default Responses workflows, you can pass attachments directly at query time with
78	`files=` and `images=`:
79
80	```python
81	from chatsnack import Chat
82
83	chat = Chat("Review the attachment and answer tersely.")
84	print(chat.ask("Summarize these.", files=["./images/chart.png", "./data/sales.csv"]))
85
86	thread = chat.chat("What stands out in this chart?", files=["./images/chart.png"])
87	thread = thread.chat("Now compare that with this report.", files=["./data/q2-report.pdf"])
88	print(thread.last)
89	```
90
91	When attachments are present, chatsnack stores the user turn as an expanded YAML block
92	(`text` + `files` / `i

People who liked this also viewed

6 picks

Type

chatsnack

29stars2forks

View on GitHub

How to install chatsnack?

mattie/chatsnack

$git clone https://github.com/Mattie/chatsnack

Installs into the current project.

Summary

Files · 1

View on GitHub

README.md

1	# chatsnack
2
3	chatsnack is the easiest Python library for rapid development with OpenAI's ChatGPT API. It provides an intuitive interface for creating and managing chat-based prompts and responses, making it convenient to build complex, interactive conversations with AI.
4
5	[Documentation site](https://mattie.github.io/chatsnack/)
6
7	![chatsnack features](/docs/chatsnack_features_smaller.jpg)
8	## Setup
9
10	### Got snack?
11
12	Install the `chatsnack` package from PyPI:
13
14	```bash
15	# requires Python 3.11+
16	pip install chatsnack
17	```
18
19	### Got keys?
20
21	Add your OpenAI API key to your .env file. If you don't have a .env file, the library will create a new one for you in the local directory.
22
23	### Learn More!
24	Read more below, watch the [intro video](https://www.youtube.com/watch?v=Yjwi54rHrhw) or check out the [Getting Started notebook](notebooks/).
25
26	### Responses API Support
27	`Chat()` now defaults to the Responses family over WebSocket with `session="inherit"`.
28	If you stay on that default path--or explicitly choose another Responses transport--
29	chatsnack requires the OpenAI Python client to be `openai>=2.29.0` (or newer).
30	For explicit WebSocket support outside chatsnack's packaged dependencies, install
31	`openai[realtime]>=2.29.0`.
32
33	## Usage
34
35	### Enjoy a Quick Snack
36
37	Easiest way to get going with `chatsnack` is with built-in snack packs. Each pack is a singleton ready to mingleton.
38
39	```python
40	>>> from chatsnack.packs import ChatsnackHelp
41	>>> ChatsnackHelp.ask("What is your primary directive?")
42	```
43	> *"My primary directive is to assist users of the chatsnack Python module by answering questions
44	and helping with any problems or concerns related to the module. I aim to provide helpful and
45	informative responses based on the chatsnack module's documentation and best practices."*
46
47	You can try out other example snack packs like `Confectioner`, `Jolly`, `Chester`, `Jane`, or `Data`. (Eventually there will be an easy way to create and share your own.)
48
49	Instead of `.ask()` we can call `.chat()` which will allow us to continue a conversation.
50	```python
51	>>> mychat = ChatsnackHelp.chat("What is chatsnack?") # submits and returns a new chat object
52	>>> print(mychat.response)
53	```
54	> Chatsnack is a Python module that provides a simple and powerful interface for creating conversational agents and tools using OpenAI's ChatGPT language models. It allows you to easily build chat prompts, manage conversation flow, and integrate with ChatGPT to generate responses. With Chatsnack, you can create chatbots, AI-assisted tools, and other conversational applications using a convenient and flexible API.
55
56	Now we can add more messages to that chat however we'd like:
57	```python
58	>>> mychat.user("Respond in only six word sentences from now on.")
59	>>> mychat.asst("I promise I will do so.")
60	>>> mychat.user("How should I spend my day?")
61	>>> mychat.ask()
62	```
63	> "Explore hobbies, exercise, connect with friends."
64
65	If you want a super simple interactive conversation with you and a chatbot, you could do something like this:
66	```python
67	from chatsnack.packs import Jolly
68	yourchat = Jolly # interview a green giant
69	while (user_input := input("Chat with the bot: ")):
70	print(f"USER: {user_input}")
71	yourchat = yourchat.chat(user_input)
72	print(f"THEM: {yourchat.last}")
73	```
74
75	### Natural Attachments (Phase 3 style ergonomics)
76
77	For the default Responses workflows, you can pass attachments directly at query time with
78	`files=` and `images=`:
79
80	```python
81	from chatsnack import Chat
82
83	chat = Chat("Review the attachment and answer tersely.")
84	print(chat.ask("Summarize these.", files=["./images/chart.png", "./data/sales.csv"]))
85
86	thread = chat.chat("What stands out in this chart?", files=["./images/chart.png"])
87	thread = thread.chat("Now compare that with this report.", files=["./data/q2-report.pdf"])
88	print(thread.last)
89	```
90
91	When attachments are present, chatsnack stores the user turn as an expanded YAML block
92	(`text` + `files` / `i

People who liked this also viewed

6 picks

Type