Kilo Code

https://kilo.ai/docs
Admin
Kilo Code is an open-source VS Code extension that accelerates development with AI-driven code...
Tokens:280,560
Snippets:2,834
Trust Score:8.2
Update:2 weeks ago
Show doc for...
Tokens:
Raw
### Example workflow: start and resume a session

Source: https://kilo.ai/docs/code-with-ai/platforms/cli

Example workflow showing how to start a session, work on a task, exit with /exit, and later resume the same session with kilo --continue. The conversation history is restored.

```bash
# Start a session
kilo
# > "Create a REST API"
# ... work on the task ...
# Exit with /exit

# Later, resume the same session
kilo --continue
# Conversation history is restored, ready to continue
```

--------------------------------

### Project setup workflow steps

Source: https://kilo.ai/docs/customize/workflows

Common workflow pattern for project setup. Steps include cloning a repository template, installing dependencies, configuring environment files, initializing database/services, and running initial tests.

```markdown
1. Clone repository template
2. Install dependencies (`npm install`, `pip install -r requirements.txt`)
3. Configure environment files
4. Initialize database/services
5. Run initial tests
```

--------------------------------

### Start Kilo TUI and check version/help

Source: https://kilo.ai/docs/code-with-ai/platforms/cli

Basic Kilo CLI commands: start the TUI, check the version, and get help.

```bash
kilo
```

```bash
kilo --version
```

```bash
kilo --help
```

--------------------------------

### macOS/Linux setup script using WORKTREE_PATH

Source: https://kilo.ai/docs/automate/agent-manager

Example setup script for macOS/Linux that changes to the new worktree directory and installs dependencies. Uses the injected WORKTREE_PATH variable and runs with sh; executable permissions are not required.

```shell
#!/bin/sh
set -e

cd "$WORKTREE_PATH"
npm install
```

--------------------------------

### Example of instruction sources in kilo.jsonc

Source: https://kilo.ai/docs/customize/custom-instructions

Shows how to configure the 'instructions' key in kilo.jsonc to point to local files, globs, or URLs. URL-based sources are fetched at session start with a 5-second timeout; unreachable URLs are silently skipped.

```yaml
instructions:
  - ./docs/coding-standards.md
  - ./teams/frontend-rules.md
  - https://example.com/team-instructions.md

```

--------------------------------

### Create .kilo/run-script for macOS/Linux

Source: https://kilo.ai/docs/automate/agent-manager

Example run script for macOS/Linux that starts the dev server with npm run dev. Place it in .kilo/run-script.

```shell
#!/bin/sh
npm run dev
```

--------------------------------

### Start the local backend with bun dev serve

Source: https://kilo.ai/docs/contributing/development-environment

Starts the local backend for backend/API validation. Refer to the root TESTING.md guide for curl-based validation against the local server.

```bash
bun dev serve

```

--------------------------------

### Install OpenAI SDK for Python

Source: https://kilo.ai/docs/gateway/sdks-and-frameworks

Install the OpenAI Python package to use with the Kilo Gateway.

```bash
pip install openai
```

--------------------------------

### Install dependencies

Source: https://kilo.ai/docs/gateway/quickstart

Installs the Vercel AI SDK, the OpenAI provider package, and dotenv for environment variable loading.

```bash
npm install ai @ai-sdk/openai dotenv

```

--------------------------------

### Launch extension with a sample workspace

Source: https://kilo.ai/docs/contributing/development-environment

Examples of launching the extension against a sample workspace using the --workspace option or the final positional argument.

```bash
bun run extension --workspace ../sample-project
bun run extension:isolated -- ../sample-project
bun run extension:isolated:clean -- ../sample-project

```

--------------------------------

### Config file example (kilo.jsonc)

Source: https://kilo.ai/docs/customize/custom-modes

JSON configuration defining agents with permissions and prompts.

```json
{
  "agent": {
    "docs-writer": {
      "description": "Specialized for writing and editing technical documentation",
      "mode": "primary",
      "color": "#10B981",
      "prompt": "You are a technical writer specializing in clear documentation.",
      "permission": {
        "edit": { "*.md": "allow", "*": "deny" },
        "bash": "deny",
      },
    },
    "test-engineer": {
      "description": "Focused on writing and maintaining test suites",
      "mode": "primary",
      "prompt": "You are a test engineer focused on code quality.",
      "permission": {
        "edit": { "*.{test,spec}.{js,ts}": "allow", "*": "deny" },
      },
    },
  },
}

```

--------------------------------

### Example .git/config for project name parsing

Source: https://kilo.ai/docs/collaborate/teams/analytics

Example of a .git/config file showing the remote named 'origin' from which the project name is automatically parsed. The project name would be 'example-repo'.

```ini
[remote "origin"]
    url = git@github.com:example-co/example-repo.git
    fetch = +refs/heads/*:refs/remotes/origin/*
```

--------------------------------

### Start Agent Manager worktree sessions with multiple tasks

Source: https://kilo.ai/docs/automate/agent-manager

Start two independent Agent Manager sessions in isolated worktrees with custom branches and prompts. Each session gets its own checkout, branch, directory, and terminal. Use this for separate tasks or alternate versions of the same task to compare approaches.

```json
{
  "mode": "worktree",
  "versions": false,
  "tasks": [
    {
      "name": "API audit",
      "branchName": "agent/api-audit",
      "prompt": "Audit the API routes and report authentication risks."
    },
    {
      "name": "UI tests",
      "branchName": "agent/ui-tests",
      "prompt": "Add focused tests for the UI state transitions."
    }
  ]
}
```

--------------------------------

### kilo daemon start help output

Source: https://kilo.ai/docs/code-with-ai/platforms/cli-reference

Shows the help output for the kilo daemon start command, with the same options as the parent daemon command.

```text
start the local kilo daemon

Options:
      --help         Show help  [boolean]
      --version      Show version number  [boolean]
      --port         port to listen on  [number] [default: 0]
      --hostname     hostname to listen on  [string] [default: "127.0.0.1"]
      --mdns         enable mDNS service discovery (defaults hostname to 0.0.0.0)  [boolean] [default: false]
      --mdns-domain  custom domain name for mDNS service (default: kilo.local)  [string] [default: "kilo.local"]
      --cors         additional domains to allow for CORS  [array] [default: []]
      --json         print daemon details as JSON  [boolean]
  -f, --foreground   keep the command active until interrupted  [boolean]

```

--------------------------------

### List models response

Source: https://kilo.ai/docs/gateway/api-reference

Example response from the GET /models endpoint, showing an OpenAI-compatible model list with fields like id, object, created, owned_by, name, context_length, and pricing.

```json
{
  "data": [
    {
      "id": "anthropic/claude-sonnet-4.5",
      "object": "model",
      "created": 1739000000,
      "owned_by": "anthropic",
      "name": "Claude Sonnet 4.5",
      "context_length": 200000,
      "pricing": {
        "prompt": "0.000003",
        "completion": "0.000015"
      }
    }
  ]
}
```

--------------------------------

### STDIO Server Implementation with StdioServerTransport

Source: https://kilo.ai/docs/automate/mcp/server-transports

Example of creating an MCP server using the SDK and attaching the STDIO transport. Imports Server and StdioServerTransport, creates a server instance, and calls transport.listen() to start listening.

```javascript
import { Server } from "@modelcontextprotocol/sdk/server/index.js"
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"

const server = new Server({ name: "local-server", version: "1.0.0" })
// Register tools...

// Use STDIO transport
const transport = new StdioServerTransport(server)
transport.listen()
```

--------------------------------

### kilo debug startup command

Source: https://kilo.ai/docs/code-with-ai/platforms/cli-reference

Prints startup timing. No positional arguments are required.

```text
kilo debug startup
```

--------------------------------

### Verify Kilo CLI Installation

Source: https://kilo.ai/docs/code-with-ai/platforms/cli

Verifies the Kilo CLI installation by printing the installed version. Run this after installation to confirm the CLI is working.

```bash
kilo --version
```

--------------------------------

### Create project directory

Source: https://kilo.ai/docs/gateway/quickstart

Creates a new project directory and initializes a Node.js project with npm.

```bash
mkdir my-ai-app
cd my-ai-app
npm init -y

```

--------------------------------

### Install Git on macOS with Xcode Command Line Tools

Source: https://kilo.ai/docs/code-with-ai/features/checkpoints

Alternative macOS installation that installs Git via Xcode Command Line Tools.

```bash
xcode-select --install
```

--------------------------------

### kilo github install command

Source: https://kilo.ai/docs/code-with-ai/platforms/cli-reference

Installs the GitHub agent.

```text
install the GitHub agent

Options:
  --help     Show help  [boolean]
  --version  Show version number  [boolean]
```

--------------------------------

### Permission configuration example

Source: https://kilo.ai/docs/customize/custom-modes

Example of permission rules using glob patterns and actions allow, deny, and ask.

```yaml
permission:
  edit:
    "*.md": "allow"
    "*": "deny"
  bash: deny
  read: allow
```

--------------------------------

### Install Kilo CLI via npm

Source: https://kilo.ai/docs/code-with-ai/platforms/cli

Installs the Kilo Code CLI globally via npm. Requires Node.js and npm. After installation, the kilo command becomes available in the terminal.

```bash
npm install -g @kilocode/cli
```

--------------------------------

### Start Kilo server with authentication

Source: https://kilo.ai/docs/code-with-ai/agents/session-history

Starts a local Kilo server with authentication on port 4096. Set the KILO_SERVER_PASSWORD environment variable to a strong password before running the server.

```bash
export KILO_SERVER_PASSWORD='replace-with-a-strong-password'
kilo serve --port 4096

```

--------------------------------

### Start a Cloud Agent task with kilo cloud start

Source: https://kilo.ai/docs/code-with-ai/platforms/cloud-agent

Starts a Cloud Agent task from the CLI. The prompt and repository are required; repository, branch, model, mode, and organization are inferred from the local checkout and CLI defaults unless matching flags are passed.

```bash
kilo cloud start --prompt "Fix the flaky login test" --repo Kilo-Org/kilocode
```

--------------------------------

### Install Kilo Code VS Code Extension and CLI

Source: https://kilo.ai/docs

Install the Kilo Code VS Code extension using the code CLI or install the Kilo CLI globally via npm.

```bash
# Install Kilo Code VS Code Extension
$ code --install-extension kilocode.kilo-code

# Or install via CLI
$ npm install -g @kilocode/cli
```
Tokens:
Raw