# Chalice-A4AB
[](https://github.com/ShotaOki/ChaliceA4AB/actions/workflows/test.yml)
## What is this?
Chalice plugin: Support `Agents for Amazon Bedrock`
## URL Links
**PYPI :: Chalice-a4ab**
https://pypi.org/project/chalice-a4ab/
**Github :: Chalice-a4ab**
https://github.com/ShotaOki/ChaliceA4AB
## Usage
1. Install
```
pip install -U chalice chalice-a4ab pydantic
```
1. Replace `from chalice import Chalice` to `from chalice_a4ab import Chalice`.
Before:
```python
from chalice import Chalice
app = Chalice("app-name")
@app.router("path-name")
...
```
After:
```python
from chalice_a4ab import Chalice
app = Chalice("app-name")
@app.router("path-name")
...
```
1. Update requirements.txt
```txt
chalice-a4ab
pydantic
```
1. Read parameters in function
```python
from chalice_a4ab import read_session_attributes, read_prompt_session_attributes, read_body
@app.router("path-name")
def method_name():
# Read body
body: PyDanticParserClass = read_body(app, PyDanticParserClass)
# Read session attributes :: key -> session attribute key
session_attributes: dict = read_session_attributes(app, "KEY")
# Read prompt session attributes :: key -> prompt session attribute key
prompt_session_attributes: dict = read_prompt_session_attributes(app, "KEY")
# Return response (dict type)
return {
"response": "object"
}
```
or read directory
```python
@app.router("path-name")
def method_name():
# Read body
body : dict = app.current_request.json_body
# Read Session attributes or Prompt attributes from header
for k in app.current_request.headers:
if k.startsWith("sessionAttributes")
# Session attributes
print(app.current_request.headers[k])
if k.startsWith("promptSessionAttributes")
# Prompt session attributes
print(app.current_request.headers[k])
return {}
```
1. **(Optional)** Add Parser Lambda Function
Add these decorator, function becomes "parser lambda"
```python
from chalice_a4ab import (
Chalice,
AgentsForAmazonBedrockConfig,
ParserLambdaAbortException,
ParserLambdaResponseModel,
)
# Set Config for Agents for Amazon bedrock
AgentsForAmazonBedrockConfig().apply()
...
# PRE_PROCESSING
@app.parser_lambda_pre_processing()
def pre_processing(event, default_result: ParserLambdaResponseModel) -> ParserLambdaResponseModel:
# [MEMO] is_valid_input :: Allow/Deny API invocation
# default_result.pre_processing_parsed_response.is_valid_input = True
return default_result
# ORCHESTRATION
@app.parser_lambda_orchestration()
def orchestration(event: dict, default_result: ParserLambdaResponseModel) -> ParserLambdaResponseModel:
# [MEMO] Throw this Exception, Overwrite LLM response
# raise ParserLambdaAbortException(message="Overwrited")
return default_result
```
1. Application works ::
| What you can do | Status |
| -------------------------------------- | ------ |
| Execute from Agents for Amazon Bedrock | 〇 |
| Auto generate OpenAPI schema | × |
| Management on cli | × |
## Advanced Usage
Create OpenAPI Schema automatically.
1. Install Chalice-a4ab and Chalice-spec
```python
pip install -U chalice chalice-spec==0.7.0 chalice-a4ab boto3 pydantic
```
1. Update requirements.txt
```txt
chalice-spec==0.7.0
chalice-a4ab
pydantic
```
1. Add Setting
```python
from chalice_a4ab import Chalice, AgentsForAmazonBedrockConfig, ModelTypesAntropicClaudeV2
from chalice_spec.docs import Docs, Operation
# Set Config for Agents for Amazon bedrock
AgentsForAmazonBedrockConfig(
# Agent Situation
instructions="Situation Settings for talking with Human and agent.(more than 40 words)",
# Agent Description
description="Description of application",
# Use Model :: Claude V2
foundation_model=ModelTypesAntropicClaudeV2,
).apply()
app = Chalice(app_name="app-name")
@app.router("path-name",
methods=["POST"],
docs=Docs(
post=Operation(
request=PyDanticRequestModelClass,
response=PyDanticOutputModelClass,
)
))
def post_method():
"""
Function summary
Description of this method, brabrarba, this comment becomes LLM Prompt.
"""
...
```
1. Read values with PyDantic
```python
@app.router("path-name",
methods=["POST"],
docs=Docs(
post=Operation(
request=PyDanticRequestModelClass,
response=PyDanticOutputModelClass,
)
))
def method_name():
"""
Function summary
Description of this method, brabrarba, this comment becomes LLM Prompt.
"""
# Read body with utility
body: PyDanticRequestModelClass = read_body(app, PyDanticRequestModelClass)
# or direct
# Pydantic V1 -> parse_obj()
# Pydantic V2 -> model_validate()
body = PyDanticRequestModelClass.model_validate(app.current_request.json_body)
# return response as dict class
# Pydantic V1 -> json()
# Pydantic V2 -> model_dump_json()
return PyDanticOutputModelClass(
response=...
).model_dump_json()
```
documentation for `@app.router` sample: https://github.com/TestBoxLab/chalice-spec
1. Management CLI
**Init command** :: Create AWS Resource and OpenAPI Schema.
```bash
chalice-a4ab init --profile ${PROFILE_NAME} --region ${REGION_NAME}
```
**Sync command** :: Update Already AWS Resource and OpenAPISchema.
```bash
chalice-a4ab sync --profile ${PROFILE_NAME} --region ${REGION_NAME}
```
**Show command** :: Show OpenAPI Schema.
```bash
chalice-a4ab show --profile ${PROFILE_NAME} --region ${REGION_NAME}
```
**Info command** :: Show about Agents for Amazon Bedrock.
```bash
chalice-a4ab info --profile ${PROFILE_NAME} --region ${REGION_NAME}
```
**Invoke command** :: Invoke Agents for Amazon Bedrock.
```bash
chalice-a4ab invoke "Natural Language Query" \
--agent-id ${AGENT_ID} \
--agent-alias-id ${AGENT_ALIAS_ID} \
--end-session \
--profile ${PROFILE_NAME} --region ${REGION_NAME}
```
**Delete AWS Resource**
```bash
chalice-a4ab delete --profile ${PROFILE_NAME} --region ${REGION_NAME}
```
1. Application works ::
| What you can do | Status |
| -------------------------------------- | ------ |
| Execute from Agents for Amazon Bedrock | 〇 |
| Auto generate OpenAPI schema | 〇 |
| Management on cli | 〇 |
# Develop
1. Setup
```bash
poetry install
```
1. Run test
```bash
poetry run pytest
```
# Lisence
MIT
# API
## Command Line TOOL
| Command | Descritpion |
| :------------------ | :------------------------------------------------ |
| chalice-a4ab init | Create AWS resource for Agents for amazon bedrock |
| chalice-a4ab sync | Sync OpenAPI schema to AWS |
| chalice-a4ab delete | Delete AWS resource for Agents for amazon bedrock |
| Options | Description |
| :-------- | :------------------------------------------ |
| --bucket | Set S3 bucket name (for put OpenAPI schema) |
| --profile | Set AWS Profile Name |
| --region | Set AWS Region Name |
| --help | Show Help |
## API
**AgentsForAmazonBedrockConfig**
| Method | Type | Description |
| :----------------------------- | :----- | :--------------------------------------- |
| apply | - | Current instace becomes global variable. |
| agents_for_bedrock_schema_json | - | Get OpenAPI Schema |
| save_schema_to_local | - | Save OpenAPI Schema to local folder |
| save_schema_to_s3 | - | Upload OpenAPI Schema to S3 bucket |
| save_config_to_local | - | Save Config setting to local folder |
| get_global_config | static | Get global variable. |
Raw data
{
"_id": null,
"home_page": "https://github.com/ShotaOki/ChaliceA4AB",
"name": "chalice-a4ab",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8,<4.0",
"maintainer_email": "",
"keywords": "Chalice,AWS,Agents for Amazon Bedrock,Bedrock",
"author": "Shota Oki",
"author_email": "",
"download_url": "https://files.pythonhosted.org/packages/f2/6d/72775c7bfd92259e1290327f4b29132d012b8bbc0f5d8e15641b7e1f7bf9/chalice_a4ab-0.3.1.tar.gz",
"platform": null,
"description": "# Chalice-A4AB\n\n\n\n[](https://github.com/ShotaOki/ChaliceA4AB/actions/workflows/test.yml)\n\n\n\n\n## What is this?\n\nChalice plugin: Support `Agents for Amazon Bedrock`\n\n## URL Links\n\n**PYPI :: Chalice-a4ab** \nhttps://pypi.org/project/chalice-a4ab/\n\n**Github :: Chalice-a4ab** \nhttps://github.com/ShotaOki/ChaliceA4AB\n\n## Usage\n\n1. Install\n\n ```\n pip install -U chalice chalice-a4ab pydantic\n ```\n\n1. Replace `from chalice import Chalice` to `from chalice_a4ab import Chalice`.\n\n Before:\n\n ```python\n from chalice import Chalice\n\n app = Chalice(\"app-name\")\n\n @app.router(\"path-name\")\n ...\n ```\n\n After:\n\n ```python\n from chalice_a4ab import Chalice\n\n app = Chalice(\"app-name\")\n\n @app.router(\"path-name\")\n ...\n ```\n\n1. Update requirements.txt\n\n ```txt\n chalice-a4ab\n pydantic\n ```\n\n1. Read parameters in function\n\n ```python\n from chalice_a4ab import read_session_attributes, read_prompt_session_attributes, read_body\n\n @app.router(\"path-name\")\n def method_name():\n\n # Read body\n body: PyDanticParserClass = read_body(app, PyDanticParserClass)\n\n # Read session attributes :: key -> session attribute key\n session_attributes: dict = read_session_attributes(app, \"KEY\")\n\n # Read prompt session attributes :: key -> prompt session attribute key\n prompt_session_attributes: dict = read_prompt_session_attributes(app, \"KEY\")\n\n # Return response (dict type)\n return {\n \"response\": \"object\"\n }\n ```\n\n or read directory\n\n ```python\n @app.router(\"path-name\")\n def method_name():\n # Read body\n body : dict = app.current_request.json_body\n\n # Read Session attributes or Prompt attributes from header\n for k in app.current_request.headers:\n if k.startsWith(\"sessionAttributes\")\n # Session attributes\n print(app.current_request.headers[k])\n if k.startsWith(\"promptSessionAttributes\")\n # Prompt session attributes\n print(app.current_request.headers[k])\n\n return {}\n ```\n\n1. **(Optional)** Add Parser Lambda Function\n\n Add these decorator, function becomes \"parser lambda\"\n\n ```python\n from chalice_a4ab import (\n Chalice,\n AgentsForAmazonBedrockConfig,\n ParserLambdaAbortException,\n ParserLambdaResponseModel,\n )\n\n # Set Config for Agents for Amazon bedrock\n AgentsForAmazonBedrockConfig().apply()\n\n ...\n\n # PRE_PROCESSING\n @app.parser_lambda_pre_processing()\n def pre_processing(event, default_result: ParserLambdaResponseModel) -> ParserLambdaResponseModel:\n # [MEMO] is_valid_input :: Allow/Deny API invocation\n # default_result.pre_processing_parsed_response.is_valid_input = True\n return default_result\n\n\n # ORCHESTRATION\n @app.parser_lambda_orchestration()\n def orchestration(event: dict, default_result: ParserLambdaResponseModel) -> ParserLambdaResponseModel:\n # [MEMO] Throw this Exception, Overwrite LLM response\n # raise ParserLambdaAbortException(message=\"Overwrited\")\n return default_result\n ```\n\n1. Application works ::\n\n | What you can do | Status |\n | -------------------------------------- | ------ |\n | Execute from Agents for Amazon Bedrock | \u3007 |\n | Auto generate OpenAPI schema | \u00d7 |\n | Management on cli | \u00d7 |\n\n## Advanced Usage\n\nCreate OpenAPI Schema automatically.\n\n1. Install Chalice-a4ab and Chalice-spec\n\n ```python\n pip install -U chalice chalice-spec==0.7.0 chalice-a4ab boto3 pydantic\n ```\n\n1. Update requirements.txt\n\n ```txt\n chalice-spec==0.7.0\n chalice-a4ab\n pydantic\n ```\n\n1. Add Setting\n\n ```python\n from chalice_a4ab import Chalice, AgentsForAmazonBedrockConfig, ModelTypesAntropicClaudeV2\n from chalice_spec.docs import Docs, Operation\n\n # Set Config for Agents for Amazon bedrock\n AgentsForAmazonBedrockConfig(\n # Agent Situation\n instructions=\"Situation Settings for talking with Human and agent.(more than 40 words)\",\n # Agent Description\n description=\"Description of application\",\n # Use Model :: Claude V2\n foundation_model=ModelTypesAntropicClaudeV2,\n ).apply()\n\n app = Chalice(app_name=\"app-name\")\n\n @app.router(\"path-name\",\n methods=[\"POST\"],\n docs=Docs(\n post=Operation(\n request=PyDanticRequestModelClass,\n response=PyDanticOutputModelClass,\n )\n ))\n def post_method():\n \"\"\"\n Function summary\n\n Description of this method, brabrarba, this comment becomes LLM Prompt.\n \"\"\"\n ...\n ```\n\n1. Read values with PyDantic\n\n ```python\n @app.router(\"path-name\",\n methods=[\"POST\"],\n docs=Docs(\n post=Operation(\n request=PyDanticRequestModelClass,\n response=PyDanticOutputModelClass,\n )\n ))\n def method_name():\n \"\"\"\n Function summary\n\n Description of this method, brabrarba, this comment becomes LLM Prompt.\n \"\"\"\n\n # Read body with utility\n body: PyDanticRequestModelClass = read_body(app, PyDanticRequestModelClass)\n\n # or direct\n # Pydantic V1 -> parse_obj()\n # Pydantic V2 -> model_validate()\n body = PyDanticRequestModelClass.model_validate(app.current_request.json_body)\n\n # return response as dict class\n # Pydantic V1 -> json()\n # Pydantic V2 -> model_dump_json()\n return PyDanticOutputModelClass(\n response=...\n ).model_dump_json()\n ```\n\n documentation for `@app.router` sample: https://github.com/TestBoxLab/chalice-spec\n\n1. Management CLI\n\n **Init command** :: Create AWS Resource and OpenAPI Schema.\n\n ```bash\n chalice-a4ab init --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n **Sync command** :: Update Already AWS Resource and OpenAPISchema.\n\n ```bash\n chalice-a4ab sync --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n **Show command** :: Show OpenAPI Schema.\n\n ```bash\n chalice-a4ab show --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n **Info command** :: Show about Agents for Amazon Bedrock.\n\n ```bash\n chalice-a4ab info --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n **Invoke command** :: Invoke Agents for Amazon Bedrock.\n\n ```bash\n chalice-a4ab invoke \"Natural Language Query\" \\\n --agent-id ${AGENT_ID} \\\n --agent-alias-id ${AGENT_ALIAS_ID} \\\n --end-session \\\n --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n **Delete AWS Resource**\n\n ```bash\n chalice-a4ab delete --profile ${PROFILE_NAME} --region ${REGION_NAME}\n ```\n\n1. Application works ::\n\n | What you can do | Status |\n | -------------------------------------- | ------ |\n | Execute from Agents for Amazon Bedrock | \u3007 |\n | Auto generate OpenAPI schema | \u3007 |\n | Management on cli | \u3007 |\n\n# Develop\n\n1. Setup\n\n ```bash\n poetry install\n ```\n\n1. Run test\n\n ```bash\n poetry run pytest\n ```\n\n# Lisence\n\nMIT\n\n# API\n\n## Command Line TOOL\n\n| Command | Descritpion |\n| :------------------ | :------------------------------------------------ |\n| chalice-a4ab init | Create AWS resource for Agents for amazon bedrock |\n| chalice-a4ab sync | Sync OpenAPI schema to AWS |\n| chalice-a4ab delete | Delete AWS resource for Agents for amazon bedrock |\n\n| Options | Description |\n| :-------- | :------------------------------------------ |\n| --bucket | Set S3 bucket name (for put OpenAPI schema) |\n| --profile | Set AWS Profile Name |\n| --region | Set AWS Region Name |\n| --help | Show Help |\n\n## API\n\n**AgentsForAmazonBedrockConfig**\n\n| Method | Type | Description |\n| :----------------------------- | :----- | :--------------------------------------- |\n| apply | - | Current instace becomes global variable. |\n| agents_for_bedrock_schema_json | - | Get OpenAPI Schema |\n| save_schema_to_local | - | Save OpenAPI Schema to local folder |\n| save_schema_to_s3 | - | Upload OpenAPI Schema to S3 bucket |\n| save_config_to_local | - | Save Config setting to local folder |\n| get_global_config | static | Get global variable. |\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Chalice x Agents for Amazon Bedrock plug-ins",
"version": "0.3.1",
"project_urls": {
"Homepage": "https://github.com/ShotaOki/ChaliceA4AB",
"Repository": "https://github.com/ShotaOki/ChaliceA4AB"
},
"split_keywords": [
"chalice",
"aws",
"agents for amazon bedrock",
"bedrock"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "e5fc3333a5b4807206287cb45221ad13d45622853fb73a50649bc63923c1cf6d",
"md5": "d815b331290af7ba52e8262706d75d3e",
"sha256": "c3772f0f65acc3e05326d5f2040567cb37b2f40469729fe290e4db25f8a4aef4"
},
"downloads": -1,
"filename": "chalice_a4ab-0.3.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "d815b331290af7ba52e8262706d75d3e",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8,<4.0",
"size": 39086,
"upload_time": "2024-01-08T02:55:19",
"upload_time_iso_8601": "2024-01-08T02:55:19.074072Z",
"url": "https://files.pythonhosted.org/packages/e5/fc/3333a5b4807206287cb45221ad13d45622853fb73a50649bc63923c1cf6d/chalice_a4ab-0.3.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "f26d72775c7bfd92259e1290327f4b29132d012b8bbc0f5d8e15641b7e1f7bf9",
"md5": "1dcf8c119f40dedffedd325d594f3714",
"sha256": "6cd10e99ab2672d4f887654e9bc660133ff6a7b319fdfd403d7770de6b8c4446"
},
"downloads": -1,
"filename": "chalice_a4ab-0.3.1.tar.gz",
"has_sig": false,
"md5_digest": "1dcf8c119f40dedffedd325d594f3714",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8,<4.0",
"size": 26545,
"upload_time": "2024-01-08T02:55:20",
"upload_time_iso_8601": "2024-01-08T02:55:20.861718Z",
"url": "https://files.pythonhosted.org/packages/f2/6d/72775c7bfd92259e1290327f4b29132d012b8bbc0f5d8e15641b7e1f7bf9/chalice_a4ab-0.3.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-01-08 02:55:20",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "ShotaOki",
"github_project": "ChaliceA4AB",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "chalice-a4ab"
}
JSON
