| Name | bridgic JSON |
| Version |
0.1.0
JSON |
| download |
| home_page | None |
| Summary | Bridgic is a project that enables developers to develop agents/workflows in a more natural and flexible way, consisting of several Python packages. |
| upload_time | 2025-10-20 12:39:34 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.9 |
| license | MIT |
| keywords |
llm
agent
devtools
workflow
|
| VCS |
|
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
Bridgic is an innovative programming framework designed to create agentic systems, from simple workflows to fully autonomous agents. Its APIs are thoughtfully crafted to be both simple and powerful.
## Core Features
* **Orchestration**: Bridgic helps to manage the control flow of your AI applications asynchronously.
* **Dynamic Control Flow**: Bridgic supports dynamic routing based on input data, and even allows workers to be added or removed at runtime.
* **Modularity**: In Bridgic, a complex agentic system can be composed by reusing components through hierarchical nesting.
* **Human-in-the-Loop**: A workflow or an agent built with Bridgic can request feedback from humans whenever needed.
* **Serialization**: Bridgic includes serialization, deserialization, and resuming capabilities to support human-in-the-loop.
* **Parameter Binding**: There are three ways to pass data among workers, including Arguments Mapping, Arguments Injection, and Inputs Propagation.
* **Systematic Integration**: A wide range of tools and LLMs can be seamlessly integrated into the Bridgic world, in a systematic way.
* **Customization**: What Bridgic provides is not a "black box" approach. You have full control over every aspect of your AI applications, such as prompts, context windows, the control flow, and more.
## Install Bridgic
Python version 3.9 or higher is required.
```bash
pip install bridgic
```
## Example Code
Initialize the running environment for LLM:
```python
import os
from bridgic.llms.openai.openai_llm import OpenAILlm, OpenAIConfiguration
_api_key = os.environ.get("OPENAI_API_KEY")
_model_name = os.environ.get("OPENAI_MODEL_NAME")
llm = OpenAILlm(
api_key=_api_key,
configuration=OpenAIConfiguration(model=_model_name),
)
```
Then, create a `word learning assistant` with code:
```python
from bridgic.core.automa import GraphAutoma, worker
from bridgic.core.model.types import Message, Role
class WordLearningAssistant(GraphAutoma):
@worker(is_start=True)
async def generate_derivatives(self, word: str):
response = await llm.achat(
model=_model_name,
messages=[
Message.from_text(text="You are a word learning assistant. Generate derivatives of the input word in a list.", role=Role.SYSTEM),
Message.from_text(text=word, role=Role.USER),
]
)
return response.message.content
@worker(dependencies=["generate_derivatives"], is_output=True)
async def make_sentences(self, derivatives):
response = await llm.achat(
model=_model_name,
messages=[
Message.from_text(text="You are a word learning assistant. Make sentences with the input derivatives in a list.", role=Role.SYSTEM),
Message.from_text(text=derivatives, role=Role.USER),
]
)
return response.message.content
```
Let's run it:
```python
word_learning_assistant = WordLearningAssistant()
res = await word_learning_assistant.arun(word="happy")
print(res)
```
For more information and examples, see the [Tutorials](https://docs.bridgic.ai/tutorials/).
## Understanding
See [Understanding Bridgic](https://docs.bridgic.ai/home/introduction/).
## License
This repo is available under the [MIT license](/LICENSE).
Raw data
{
"_id": null,
"home_page": null,
"name": "bridgic",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "LLM, agent, devtools, workflow",
"author": null,
"author_email": "Tielei Zhang <zhangtl04@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/2f/4e/9a9170f25be7c27dac732236de83acac36a0d461d5e288f3824268ede3e6/bridgic-0.1.0.tar.gz",
"platform": null,
"description": "Bridgic is an innovative programming framework designed to create agentic systems, from simple workflows to fully autonomous agents. Its APIs are thoughtfully crafted to be both simple and powerful.\n\n## Core Features\n\n* **Orchestration**: Bridgic helps to manage the control flow of your AI applications asynchronously.\n* **Dynamic Control Flow**: Bridgic supports dynamic routing based on input data, and even allows workers to be added or removed at runtime.\n* **Modularity**: In Bridgic, a complex agentic system can be composed by reusing components through hierarchical nesting.\n* **Human-in-the-Loop**: A workflow or an agent built with Bridgic can request feedback from humans whenever needed.\n* **Serialization**: Bridgic includes serialization, deserialization, and resuming capabilities to support human-in-the-loop.\n* **Parameter Binding**: There are three ways to pass data among workers, including Arguments Mapping, Arguments Injection, and Inputs Propagation.\n* **Systematic Integration**: A wide range of tools and LLMs can be seamlessly integrated into the Bridgic world, in a systematic way.\n* **Customization**: What Bridgic provides is not a \"black box\" approach. You have full control over every aspect of your AI applications, such as prompts, context windows, the control flow, and more.\n\n## Install Bridgic\n\nPython version 3.9 or higher is required.\n\n```bash\npip install bridgic\n```\n\n## Example Code\n\nInitialize the running environment for LLM:\n\n```python\nimport os\nfrom bridgic.llms.openai.openai_llm import OpenAILlm, OpenAIConfiguration\n\n_api_key = os.environ.get(\"OPENAI_API_KEY\")\n_model_name = os.environ.get(\"OPENAI_MODEL_NAME\")\n\nllm = OpenAILlm(\n api_key=_api_key,\n configuration=OpenAIConfiguration(model=_model_name),\n)\n```\n\nThen, create a `word learning assistant` with code:\n\n```python\nfrom bridgic.core.automa import GraphAutoma, worker\nfrom bridgic.core.model.types import Message, Role\n\nclass WordLearningAssistant(GraphAutoma):\n @worker(is_start=True)\n async def generate_derivatives(self, word: str):\n response = await llm.achat(\n model=_model_name,\n messages=[\n Message.from_text(text=\"You are a word learning assistant. Generate derivatives of the input word in a list.\", role=Role.SYSTEM),\n Message.from_text(text=word, role=Role.USER),\n ]\n )\n return response.message.content\n\n @worker(dependencies=[\"generate_derivatives\"], is_output=True)\n async def make_sentences(self, derivatives):\n response = await llm.achat(\n model=_model_name,\n messages=[\n Message.from_text(text=\"You are a word learning assistant. Make sentences with the input derivatives in a list.\", role=Role.SYSTEM),\n Message.from_text(text=derivatives, role=Role.USER),\n ]\n )\n return response.message.content\n```\n\nLet's run it:\n\n```python\nword_learning_assistant = WordLearningAssistant()\nres = await word_learning_assistant.arun(word=\"happy\")\nprint(res)\n```\n\nFor more information and examples, see the [Tutorials](https://docs.bridgic.ai/tutorials/).\n\n## Understanding\n\nSee [Understanding Bridgic](https://docs.bridgic.ai/home/introduction/).\n\n## License\n\nThis repo is available under the [MIT license](/LICENSE).",
"bugtrack_url": null,
"license": "MIT",
"summary": "Bridgic is a project that enables developers to develop agents/workflows in a more natural and flexible way, consisting of several Python packages.",
"version": "0.1.0",
"project_urls": null,
"split_keywords": [
"llm",
" agent",
" devtools",
" workflow"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "cfc6059d4a0753cc0b9b165e121c8ad24c1eb450c3b00519dbc2120b59286cda",
"md5": "9234782e1ff0027e9b130ddef4d8793a",
"sha256": "a71b3710bcf209382026e157d2078cd757b1af1c3e5b54828dd56770f3f61e51"
},
"downloads": -1,
"filename": "bridgic-0.1.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "9234782e1ff0027e9b130ddef4d8793a",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 3443,
"upload_time": "2025-10-20T12:39:33",
"upload_time_iso_8601": "2025-10-20T12:39:33.311354Z",
"url": "https://files.pythonhosted.org/packages/cf/c6/059d4a0753cc0b9b165e121c8ad24c1eb450c3b00519dbc2120b59286cda/bridgic-0.1.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "2f4e9a9170f25be7c27dac732236de83acac36a0d461d5e288f3824268ede3e6",
"md5": "526b86a83a87a9c820ec3f13444dc6d2",
"sha256": "38d8bd94acc85d06536073f34bede9ed3b5e1817deb2b8b9ee3451ad6b4b8e93"
},
"downloads": -1,
"filename": "bridgic-0.1.0.tar.gz",
"has_sig": false,
"md5_digest": "526b86a83a87a9c820ec3f13444dc6d2",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 3094,
"upload_time": "2025-10-20T12:39:34",
"upload_time_iso_8601": "2025-10-20T12:39:34.368193Z",
"url": "https://files.pythonhosted.org/packages/2f/4e/9a9170f25be7c27dac732236de83acac36a0d461d5e288f3824268ede3e6/bridgic-0.1.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-10-20 12:39:34",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "bridgic"
}
