Mistral

Mistral Agents adapter for Flyte (mistralai 2.x).

Bring your own Mistral agent and run it durably on Flyte. Tools you expose are Flyte tasks (each call a durable child action), and each model turn is recorded via flyte.trace (the turns are in-process HTTP calls, so we can trace the seam below the SDK’s loop) for per-turn replay on retry.

• flyteplugins.agents.mistral.tool — turn an @env.task into a Mistral run-framework tool.
• flyteplugins.agents.mistral.run_agent — run the SDK’s agent loop inside your task; return the answer.

Directory

Methods

Methods

run_agent()

def run_agent(
    input: str,
    tools: typing.Sequence[typing.Any] = (),
    model: str | None = 'mistral-large-latest',
    instructions: str | None = None,
    timeout_ms: int | None = None,
    durable: bool = True,
    observability: bool = True,
    agent_id: str | None = None,
    api_key_env_var: str = 'MISTRAL_API_KEY',
    memory_key: str | None = None,
) -> str

Run a Mistral agent with the given tools and prompt; return the final text.

Await this from an async task as await run_agent(...); from a sync task use flyteplugins.agents.mistral.run_agent_sync instead.

Call this from inside an @env.task — that task is the durable parent. The Mistral SDK runs the agent loop; each tool the agent calls runs as a durable Flyte child action, and (with durable=True) each model turn is recorded for replay. Pass agent_id to drive a pre-created server-side agent instead of an inline model.

run_agent_sync()

def run_agent_sync(
    input: str,
    tools: typing.Sequence[typing.Any] = (),
    model: str | None = 'mistral-large-latest',
    instructions: str | None = None,
    timeout_ms: int | None = None,
    durable: bool = True,
    observability: bool = True,
    agent_id: str | None = None,
    api_key_env_var: str = 'MISTRAL_API_KEY',
    memory_key: str | None = None,
) -> str

Synchronous variant of run_agent for use in sync tasks; runs the async implementation on a dedicated event loop.

Run a Mistral agent with the given tools and prompt; return the final text.

Await this from an async task as await run_agent(...); from a sync task use flyteplugins.agents.mistral.run_agent_sync instead.

Call this from inside an @env.task — that task is the durable parent. The Mistral SDK runs the agent loop; each tool the agent calls runs as a durable Flyte child action, and (with durable=True) each model turn is recorded for replay. Pass agent_id to drive a pre-created server-side agent instead of an inline model.

tool()

def tool(
    func: AsyncFunctionTaskTemplate | typing.Callable | None = None,
    name: str | None = None,
    description: str | None = None,
) -> typing.Callable

Wrap a Flyte @env.task as a plain async tool function — the generic default.

For SDKs that accept plain Python callables as tools (deriving the schema from the signature + docstring), this is the whole adapter tool: the returned function carries the task’s signature (functools.wraps), dispatches to task.aio() (so each call is a durable Flyte child action), exposes __wrapped_task__, and wires the backing task to flyteplugins.agents.core.ToolTaskResolver. Adapters whose SDK needs a native tool type (e.g. OpenAI’s FunctionTool, Claude’s MCP SdkMcpTool) provide their own instead.

Also accepts any other callable — a plain function or an instance of a callable class defining __call__ — and returns it usable as a tool as-is, since the plain-callable SDKs derive the schema by inspecting the callable (a class instance is inspected through its __call__). A name or description override is applied to the callable best-effort.

Usable bare, parametrized or as a direct call:

@tool
@env.task
async def get_weather(city: str) -> str: ...