# AI Mesh SDK
A lightweight Python SDK for discovering and calling agents on the AI Mesh platform.
## Features
- List available agents on the mesh
- Call any agent by ID with custom inputs
- **Input validation** using JSON Schema (when available)
- Convert all Mesh agents into LangChain-compatible tools
- Auto-generated parameter documentation
- Simple authentication with API tokens
## Installation
```bash
pip install ai-mesh-sdk
```
## Quick Start
```python
from mesh_sdk import MeshSDK
# Initialize with your API token
sdk = MeshSDK(token="your-api-token")
# List all available agents
agents = sdk.list_agents()
print(f"Found {len(agents)} agents")
# Call a specific agent
result = sdk.call_agent(
agent_id="agent-123",
inputs={"prompt": "Hello, world!"}
)
print(result)
# Use with LangChain
from langchain.agents import initialize_agent, AgentType
tools = sdk.to_langchain_tools()
agent = initialize_agent(
tools=tools,
llm=your_llm,
agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION
)
```
## API Reference
### MeshSDK(token)
Initialize the SDK with your API token.
**Parameters:**
- `token` (str): Your AI Mesh API token
### list_agents()
Returns a list of all available agents on the mesh.
**Returns:** List[Dict] - Agent metadata including ID, name, and description
### call_agent(agent_id, inputs, validate_inputs=True)
Call a specific agent with provided inputs.
**Parameters:**
- `agent_id` (str): The unique identifier of the agent
- `inputs` (Dict): Input parameters for the agent
- `validate_inputs` (bool): Whether to validate inputs against schema (default: True)
**Returns:** Dict - The agent's response
**Input Validation:**
If an agent has an `inputSchema`, the SDK will automatically validate your inputs:
```python
# This will validate inputs against the agent's schema
result = sdk.call_agent("summarizer", {
"text": "Long text to summarize...",
"max_length": 100
})
# Skip validation if needed
result = sdk.call_agent("summarizer", inputs, validate_inputs=False)
```
### to_langchain_tools()
Convert all mesh agents into LangChain Tool objects with automatic input validation.
**Returns:** List[Tool] - LangChain-compatible tools
**Enhanced Tool Descriptions:**
Tools automatically include parameter information from the agent's schema:
```python
tools = sdk.to_langchain_tools()
for tool in tools:
print(f"{tool.name}: {tool.description}")
# Output example:
# Summarizer: Summarizes long text into concise summaries
#
# Parameters: text: string (required), max_length: integer (optional)
```
## Agent Schema Format
For optimal validation, agents should include an `inputSchema` in their metadata:
```json
{
"id": "summarizer",
"name": "Text Summarizer",
"description": "Summarizes long text into concise summaries",
"inputSchema": {
"type": "object",
"properties": {
"text": {
"type": "string",
"description": "The text to summarize"
},
"max_length": {
"type": "integer",
"description": "Maximum length of summary",
"default": 100
}
},
"required": ["text"]
}
}
```
**Fallback:** If no `inputSchema` is available, the SDK will look for `exampleInputs` and include those in the tool description.
## License
MIT License
Raw data
{
"_id": null,
"home_page": "https://github.com/yourusername/ai-mesh-sdk",
"name": "ai-mesh-sdk",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": null,
"keywords": "ai, agents, mesh, langchain, tools",
"author": "Your Name",
"author_email": "your@email.com",
"download_url": "https://files.pythonhosted.org/packages/3b/23/3ea5861b1d6ab10c3559488482559a4ee6ef5093c34689c1fb3f852b9477/ai_mesh_sdk-0.1.1.tar.gz",
"platform": null,
"description": "\n# AI Mesh SDK\n\nA lightweight Python SDK for discovering and calling agents on the AI Mesh platform.\n\n## Features\n- List available agents on the mesh\n- Call any agent by ID with custom inputs\n- **Input validation** using JSON Schema (when available)\n- Convert all Mesh agents into LangChain-compatible tools\n- Auto-generated parameter documentation\n- Simple authentication with API tokens\n\n## Installation\n\n```bash\npip install ai-mesh-sdk\n```\n\n## Quick Start\n\n```python\nfrom mesh_sdk import MeshSDK\n\n# Initialize with your API token\nsdk = MeshSDK(token=\"your-api-token\")\n\n# List all available agents\nagents = sdk.list_agents()\nprint(f\"Found {len(agents)} agents\")\n\n# Call a specific agent\nresult = sdk.call_agent(\n agent_id=\"agent-123\",\n inputs={\"prompt\": \"Hello, world!\"}\n)\nprint(result)\n\n# Use with LangChain\nfrom langchain.agents import initialize_agent, AgentType\n\ntools = sdk.to_langchain_tools()\nagent = initialize_agent(\n tools=tools,\n llm=your_llm,\n agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION\n)\n```\n\n## API Reference\n\n### MeshSDK(token)\nInitialize the SDK with your API token.\n\n**Parameters:**\n- `token` (str): Your AI Mesh API token\n\n### list_agents()\nReturns a list of all available agents on the mesh.\n\n**Returns:** List[Dict] - Agent metadata including ID, name, and description\n\n### call_agent(agent_id, inputs, validate_inputs=True)\nCall a specific agent with provided inputs.\n\n**Parameters:**\n- `agent_id` (str): The unique identifier of the agent\n- `inputs` (Dict): Input parameters for the agent\n- `validate_inputs` (bool): Whether to validate inputs against schema (default: True)\n\n**Returns:** Dict - The agent's response\n\n**Input Validation:**\nIf an agent has an `inputSchema`, the SDK will automatically validate your inputs:\n\n```python\n# This will validate inputs against the agent's schema\nresult = sdk.call_agent(\"summarizer\", {\n \"text\": \"Long text to summarize...\",\n \"max_length\": 100\n})\n\n# Skip validation if needed\nresult = sdk.call_agent(\"summarizer\", inputs, validate_inputs=False)\n```\n\n### to_langchain_tools()\nConvert all mesh agents into LangChain Tool objects with automatic input validation.\n\n**Returns:** List[Tool] - LangChain-compatible tools\n\n**Enhanced Tool Descriptions:**\nTools automatically include parameter information from the agent's schema:\n\n```python\ntools = sdk.to_langchain_tools()\nfor tool in tools:\n print(f\"{tool.name}: {tool.description}\")\n\n# Output example:\n# Summarizer: Summarizes long text into concise summaries\n# \n# Parameters: text: string (required), max_length: integer (optional)\n```\n\n## Agent Schema Format\n\nFor optimal validation, agents should include an `inputSchema` in their metadata:\n\n```json\n{\n \"id\": \"summarizer\",\n \"name\": \"Text Summarizer\", \n \"description\": \"Summarizes long text into concise summaries\",\n \"inputSchema\": {\n \"type\": \"object\",\n \"properties\": {\n \"text\": {\n \"type\": \"string\",\n \"description\": \"The text to summarize\"\n },\n \"max_length\": {\n \"type\": \"integer\", \n \"description\": \"Maximum length of summary\",\n \"default\": 100\n }\n },\n \"required\": [\"text\"]\n }\n}\n```\n\n**Fallback:** If no `inputSchema` is available, the SDK will look for `exampleInputs` and include those in the tool description.\n\n## License\n\nMIT License\n",
"bugtrack_url": null,
"license": null,
"summary": "Python SDK for interacting with AI Mesh platform",
"version": "0.1.1",
"project_urls": {
"Homepage": "https://github.com/yourusername/ai-mesh-sdk"
},
"split_keywords": [
"ai",
" agents",
" mesh",
" langchain",
" tools"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "75a671529efde338e2b654032a4f6c87965f5189cd6578d07b1b84766be7e624",
"md5": "e972e9d2f986ce87f3ba6f2c7af12602",
"sha256": "04ecf214e968712f1d4b6f9f2698bfb820a109417dd686fbb45aa3b14091f0e7"
},
"downloads": -1,
"filename": "ai_mesh_sdk-0.1.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "e972e9d2f986ce87f3ba6f2c7af12602",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 3526,
"upload_time": "2025-08-02T14:24:23",
"upload_time_iso_8601": "2025-08-02T14:24:23.466358Z",
"url": "https://files.pythonhosted.org/packages/75/a6/71529efde338e2b654032a4f6c87965f5189cd6578d07b1b84766be7e624/ai_mesh_sdk-0.1.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "3b233ea5861b1d6ab10c3559488482559a4ee6ef5093c34689c1fb3f852b9477",
"md5": "c945b59c41b51cea085dd9ace3c4cee9",
"sha256": "c025ac5bd253577e8388ed86b62870f4b9b61ed2ffe88fa4032a98506cd7c9b9"
},
"downloads": -1,
"filename": "ai_mesh_sdk-0.1.1.tar.gz",
"has_sig": false,
"md5_digest": "c945b59c41b51cea085dd9ace3c4cee9",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 3564,
"upload_time": "2025-08-02T14:24:24",
"upload_time_iso_8601": "2025-08-02T14:24:24.678158Z",
"url": "https://files.pythonhosted.org/packages/3b/23/3ea5861b1d6ab10c3559488482559a4ee6ef5093c34689c1fb3f852b9477/ai_mesh_sdk-0.1.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-02 14:24:24",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "yourusername",
"github_project": "ai-mesh-sdk",
"github_not_found": true,
"lcname": "ai-mesh-sdk"
}
JSON
