Jump to content

Connect SuperML | Leeroopedia MCP: Equip your AI agents with best practices, code verification, and debugging knowledge. Powered by Leeroo — building Organizational Superintelligence. Contact us at founders@leeroo.com.

Implementation:Ollama Ollama XCmd Run

From Leeroopedia
Revision as of 13:28, 16 February 2026 by Admin (talk | contribs) (Auto-imported from implementations/Ollama_Ollama_XCmd_Run.md)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Knowledge Sources
Domains CLI, Agent Loop
Last Updated 2025-02-15 00:00 GMT

Overview

Implements the interactive agent chat loop for ollama run, orchestrating the conversation between the user, the LLM, and tools (bash, web search, web fetch).

Description

The Chat function manages a streaming conversation with the Ollama API, handling tool calls from the LLM by routing them through the tool registry and approval manager. Detects local vs. cloud models to apply different token limits for tool output truncation (4k tokens for local, 10k for cloud). Supports cloud sign-in flow by polling the Whoami endpoint. Renders tool outputs with formatting, manages thinking display, and handles word wrapping for terminal display.

Usage

Called as the main entry point for interactive AI agent sessions with tool use capabilities in the experimental ollama run mode.

Code Reference

Source Location

  • Repository: Ollama
  • File: x/cmd/run.go
  • Lines: 1-1111

Signature

type RunOptions struct {
    Model        string
    Messages     []api.Message
    WordWrap     bool
    Format       string
    System       string
    Options      map[string]any
    KeepAlive    *api.Duration
    Think        *api.ThinkValue
    HideThinking bool
    Verbose      bool
    Tools        *tools.Registry
    Approval     *agent.ApprovalManager
    YoloMode     bool
}

func Chat(ctx context.Context, opts RunOptions) (*api.Message, error)
func truncateToolOutput(output, modelName string) string

Import

import "github.com/ollama/ollama/x/cmd"

I/O Contract

Inputs

Name Type Required Description
ctx context.Context Yes Context for cancellation
opts RunOptions Yes Configuration including model name, tools, and approval manager

Outputs

Name Type Description
message *api.Message Final assistant message with content and tool calls
error error Non-nil on API or execution errors

Usage Examples

registry := tools.DefaultRegistry()
approval := agent.NewApprovalManager()

msg, err := cmd.Chat(ctx, cmd.RunOptions{
    Model:    "qwen3:latest",
    Messages: messages,
    Tools:    registry,
    Approval: approval,
    WordWrap: true,
})

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment