{"owner":"huggingface","repo":"smolagents","hasSkills":true,"hasMcp":false,"mcpConfig":null,"found":["AGENTS.md","docs/source/en/reference/agents.md"],"skills":{"AGENTS.md":"# Contributor Guidelines\n- Follow OOP principles\n- Be Pythonic: follow Python best practices and idiomatic patterns\n- Write unit tests for new functionality\n","docs/source/en/reference/agents.md":"# Agents\n\n<Tip warning={true}>\n\nSmolagents is an experimental API which is subject to change at any time. Results returned by the agents\ncan vary as the APIs or underlying models are prone to change.\n\n</Tip>\n\nTo learn more about agents and tools make sure to read the [introductory guide](../index). This page\ncontains the API docs for the underlying classes.\n\n## Agents\n\nOur agents inherit from [`MultiStepAgent`], which means they can act in multiple steps, each step consisting of one thought, then one tool call and execution. Read more in [this conceptual guide](../conceptual_guides/react).\n\nWe provide two types of agents, based on the main [`Agent`] class.\n  - [`CodeAgent`] writes its tool calls in Python code (this is the default).\n  - [`ToolCallingAgent`] writes its tool calls in JSON.\n\nBoth require arguments `model` and list of tools `tools` at initialization.\n\n### Classes of agents\n\n[[autodoc]] MultiStepAgent\n\n[[autodoc]] CodeAgent\n\n[[autodoc]] ToolCallingAgent\n\n### stream_to_gradio\n\n[[autodoc]] stream_to_gradio\n\n### GradioUI\n\n> [!TIP]\n> You must have `gradio` installed to use the UI. Please run `pip install 'smolagents[gradio]'` if it's not the case.\n\n[[autodoc]] GradioUI\n\n## Prompts\n\n[[autodoc]] smolagents.agents.PromptTemplates\n\n[[autodoc]] smolagents.agents.PlanningPromptTemplate\n\n[[autodoc]] smolagents.agents.ManagedAgentPromptTemplate\n\n[[autodoc]] smolagents.agents.FinalAnswerPromptTemplate\n\n## Memory\n\nSmolagents use memory to store information across multiple steps.\n\n[[autodoc]] smolagents.memory.AgentMemory\n"},"files":{"AGENTS.md":"# Contributor Guidelines\n- Follow OOP principles\n- Be Pythonic: follow Python best practices and idiomatic patterns\n- Write unit tests for new functionality\n","docs/source/en/reference/agents.md":"# Agents\n\n<Tip warning={true}>\n\nSmolagents is an experimental API which is subject to change at any time. Results returned by the agents\ncan vary as the APIs or underlying models are prone to change.\n\n</Tip>\n\nTo learn more about agents and tools make sure to read the [introductory guide](../index). This page\ncontains the API docs for the underlying classes.\n\n## Agents\n\nOur agents inherit from [`MultiStepAgent`], which means they can act in multiple steps, each step consisting of one thought, then one tool call and execution. Read more in [this conceptual guide](../conceptual_guides/react).\n\nWe provide two types of agents, based on the main [`Agent`] class.\n  - [`CodeAgent`] writes its tool calls in Python code (this is the default).\n  - [`ToolCallingAgent`] writes its tool calls in JSON.\n\nBoth require arguments `model` and list of tools `tools` at initialization.\n\n### Classes of agents\n\n[[autodoc]] MultiStepAgent\n\n[[autodoc]] CodeAgent\n\n[[autodoc]] ToolCallingAgent\n\n### stream_to_gradio\n\n[[autodoc]] stream_to_gradio\n\n### GradioUI\n\n> [!TIP]\n> You must have `gradio` installed to use the UI. Please run `pip install 'smolagents[gradio]'` if it's not the case.\n\n[[autodoc]] GradioUI\n\n## Prompts\n\n[[autodoc]] smolagents.agents.PromptTemplates\n\n[[autodoc]] smolagents.agents.PlanningPromptTemplate\n\n[[autodoc]] smolagents.agents.ManagedAgentPromptTemplate\n\n[[autodoc]] smolagents.agents.FinalAnswerPromptTemplate\n\n## Memory\n\nSmolagents use memory to store information across multiple steps.\n\n[[autodoc]] smolagents.memory.AgentMemory\n"},"items":[{"name":"AGENTS.md","path":"AGENTS.md","title":"AGENTS.md","content":"# Contributor Guidelines\n- Follow OOP principles\n- Be Pythonic: follow Python best practices and idiomatic patterns\n- Write unit tests for new functionality\n","category":"root","tokens":40},{"name":"agents.md","path":"docs/source/en/reference/agents.md","title":"agents.md","content":"# Agents\n\n<Tip warning={true}>\n\nSmolagents is an experimental API which is subject to change at any time. Results returned by the agents\ncan vary as the APIs or underlying models are prone to change.\n\n</Tip>\n\nTo learn more about agents and tools make sure to read the [introductory guide](../index). This page\ncontains the API docs for the underlying classes.\n\n## Agents\n\nOur agents inherit from [`MultiStepAgent`], which means they can act in multiple steps, each step consisting of one thought, then one tool call and execution. Read more in [this conceptual guide](../conceptual_guides/react).\n\nWe provide two types of agents, based on the main [`Agent`] class.\n  - [`CodeAgent`] writes its tool calls in Python code (this is the default).\n  - [`ToolCallingAgent`] writes its tool calls in JSON.\n\nBoth require arguments `model` and list of tools `tools` at initialization.\n\n### Classes of agents\n\n[[autodoc]] MultiStepAgent\n\n[[autodoc]] CodeAgent\n\n[[autodoc]] ToolCallingAgent\n\n### stream_to_gradio\n\n[[autodoc]] stream_to_gradio\n\n### GradioUI\n\n> [!TIP]\n> You must have `gradio` installed to use the UI. Please run `pip install 'smolagents[gradio]'` if it's not the case.\n\n[[autodoc]] GradioUI\n\n## Prompts\n\n[[autodoc]] smolagents.agents.PromptTemplates\n\n[[autodoc]] smolagents.agents.PlanningPromptTemplate\n\n[[autodoc]] smolagents.agents.ManagedAgentPromptTemplate\n\n[[autodoc]] smolagents.agents.FinalAnswerPromptTemplate\n\n## Memory\n\nSmolagents use memory to store information across multiple steps.\n\n[[autodoc]] smolagents.memory.AgentMemory\n","category":"docs","tokens":387}]}