@tecet/ollm 0.1.3

package/docs/UI&Settings/Configuration.md

@@ -0,0 +1,872 @@
+# Configuration Reference
+
+## Overview
+
+OLLM CLI uses a layered configuration system that allows you to customize behavior at multiple levels. Configuration can be specified through config files, environment variables, and command-line flags, with a clear precedence order.
+
+## Configuration File Locations
+
+OLLM CLI looks for configuration files in two locations:
+
+### User Configuration
+
+**Location:** `~/.ollm/config.yaml`
+
+This is your personal configuration that applies to all OLLM CLI sessions across all projects. Use this for your preferred defaults like model selection, UI preferences, and provider endpoints.
+
+**Example path:**
+
+- Linux/macOS: `/home/username/.ollm/config.yaml`
+- Windows: `C:\Users\username\.ollm\config.yaml`
+
+### Workspace Configuration
+
+**Location:** `.ollm/config.yaml` (in your project directory)
+
+This is project-specific configuration that only applies when running OLLM CLI from within that project directory. Use this for project-specific settings like custom models, tool policies, or provider configurations.
+
+**Example path:**
+
+- Linux/macOS: `/path/to/project/.ollm/config.yaml`
+- Windows: `C:\path\to\project\.ollm\config.yaml`
+
+## Configuration Precedence
+
+When multiple configuration sources are present, they are merged with the following precedence order (highest priority first):
+
+1. **CLI Flags** - Command-line arguments (e.g., `--model llama3.1`)
+2. **Environment Variables** - Environment variables (e.g., `OLLAMA_HOST`)
+3. **Workspace Config** - Project-specific `.ollm/config.yaml`
+4. **User Config** - Personal `~/.ollm/config.yaml`
+5. **System Defaults** - Built-in default values
+
+Settings from higher priority sources override settings from lower priority sources. For nested objects, the merge is performed at the field level, not the object level.
+
+### Precedence Example
+
+```yaml
+# User config (~/.ollm/config.yaml)
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+
+# Workspace config (.ollm/config.yaml)
+model:
+  default: codellama:13b
+
+# Result: workspace model overrides user model, but temperature is inherited
+# Final config:
+#   model.default: codellama:13b
+#   model.temperature: 0.7
+```
+
+### CLI Flag Precedence
+
+```bash
+# User config has model: llama3.2:3b
+# This command overrides it:
+ollm --model llama3.1:8b
+
+# Final model used: llama3.1:8b
+```
+
+### Environment Variable Precedence
+
+```bash
+# User config has provider.ollama.host: http://localhost:11434
+# This environment variable overrides it:
+export OLLAMA_HOST=http://192.168.1.100:11434
+ollm
+
+# Final host used: http://192.168.1.100:11434
+```
+
+## Configuration File Format
+
+Configuration files use YAML format. Here's the basic structure:
+
+```yaml
+# Provider configuration
+provider:
+  default: ollama
+  ollama:
+    host: http://localhost:11434
+    timeout: 30000
+
+# Model configuration
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+  maxTokens: 4096
+
+# UI configuration
+ui:
+  layout: hybrid
+  sidePanel: true
+  showGpuStats: true
+  showCost: true
+
+# Status monitoring
+status:
+  pollInterval: 5000
+  highTempThreshold: 80
+  lowVramThreshold: 512
+
+# Diff review
+review:
+  enabled: true
+  inlineThreshold: 5
+
+# Session management
+session:
+  autoSave: true
+  saveInterval: 60000
+```
+
+## Settings Reference
+
+### Provider Settings
+
+Controls which LLM provider to use and how to connect to it.
+
+| Setting                            | Type         | Default                    | Description                                                     |
+| ---------------------------------- | ------------ | -------------------------- | --------------------------------------------------------------- |
+| `provider.default`                 | string       | `"ollama"`                 | Default provider to use (`ollama`, `vllm`, `openai-compatible`) |
+| `provider.ollama.host`             | string (URI) | `"http://localhost:11434"` | Ollama server endpoint URL                                      |
+| `provider.ollama.timeout`          | number       | `30000`                    | Request timeout in milliseconds (min: 0)                        |
+| `provider.vllm.host`               | string (URI) | -                          | vLLM server endpoint URL                                        |
+| `provider.vllm.apiKey`             | string       | -                          | Optional API key for vLLM authentication                        |
+| `provider.openaiCompatible.host`   | string (URI) | -                          | OpenAI-compatible server endpoint URL                           |
+| `provider.openaiCompatible.apiKey` | string       | -                          | Optional API key for OpenAI-compatible authentication           |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+provider:
+  default: ollama
+  ollama:
+    host: http://localhost:11434
+    timeout: 30000
+```
+
+```bash
+# Environment variables
+export OLLAMA_HOST=http://localhost:11434
+export VLLM_HOST=http://localhost:8000
+export VLLM_API_KEY=your-api-key
+export OPENAI_COMPATIBLE_HOST=http://localhost:1234
+export OPENAI_COMPATIBLE_API_KEY=your-api-key
+
+# CLI flags
+ollm --provider ollama --host http://localhost:11434
+```
+
+### Model Settings
+
+Controls model selection and generation parameters.
+
+| Setting             | Type   | Default         | Description                                            |
+| ------------------- | ------ | --------------- | ------------------------------------------------------ |
+| `model.default`     | string | `"llama3.2:3b"` | Default model name to use                              |
+| `model.temperature` | number | `0.7`           | Sampling temperature (0.0-2.0, higher = more creative) |
+| `model.maxTokens`   | number | `4096`          | Maximum tokens to generate (min: 1)                    |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+  maxTokens: 4096
+```
+
+```bash
+# Environment variables
+export OLLM_DEFAULT_MODEL=llama3.1:8b
+
+# CLI flags
+ollm --model llama3.1:8b
+```
+
+### Context Settings
+
+Controls context management and VRAM-aware sizing (optional, uses defaults if not specified).
+
+| Setting                        | Type    | Default      | Description                                   |
+| ------------------------------ | ------- | ------------ | --------------------------------------------- |
+| `context.targetSize`           | number  | `8192`       | Target context size in tokens                 |
+| `context.minSize`              | number  | `2048`       | Minimum context size in tokens                |
+| `context.maxSize`              | number  | `32768`      | Maximum context size in tokens                |
+| `context.autoSize`             | boolean | `true`       | Enable automatic context sizing based on VRAM |
+| `context.vramBuffer`           | number  | `1073741824` | VRAM buffer to reserve in bytes (1GB default) |
+| `context.compressionEnabled`   | boolean | `true`       | Enable automatic context compression          |
+| `context.compressionThreshold` | number  | `0.8`        | Threshold (0-1) to trigger compression        |
+| `context.snapshotsEnabled`     | boolean | `true`       | Enable snapshots for context restoration      |
+| `context.maxSnapshots`         | number  | `5`          | Maximum number of snapshots to keep           |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+context:
+  targetSize: 8192
+  autoSize: true
+  compressionEnabled: true
+  compressionThreshold: 0.8
+```
+
+```bash
+# Environment variables
+# (Context settings are primarily configured via config file)
+
+# CLI flags
+# (Context settings are not available as CLI flags)
+```
+
+### UI Settings
+
+Controls the terminal user interface appearance and behavior.
+
+| Setting                               | Type    | Default    | Description                                         |
+| ------------------------------------- | ------- | ---------- | --------------------------------------------------- |
+| `ui.layout`                           | string  | `"hybrid"` | UI layout mode (`hybrid` or `simple`)               |
+| `ui.sidePanel`                        | boolean | `true`     | Show side panel with context info                   |
+| `ui.showGpuStats`                     | boolean | `true`     | Display GPU statistics in status bar                |
+| `ui.showCost`                         | boolean | `true`     | Display token cost estimates                        |
+| `ui.metrics.enabled`                  | boolean | `true`     | Enable performance metrics display                  |
+| `ui.metrics.compactMode`              | boolean | `false`    | Use compact metrics display                         |
+| `ui.metrics.showPromptTokens`         | boolean | `true`     | Show prompt token count                             |
+| `ui.metrics.showTTFT`                 | boolean | `true`     | Show time to first token                            |
+| `ui.metrics.showInStatusBar`          | boolean | `true`     | Display metrics in status bar                       |
+| `ui.reasoning.enabled`                | boolean | `true`     | Enable reasoning display for models that support it |
+| `ui.reasoning.maxVisibleLines`        | number  | `8`        | Maximum lines to show for reasoning (min: 1)        |
+| `ui.reasoning.autoCollapseOnComplete` | boolean | `true`     | Auto-collapse reasoning when complete               |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+ui:
+  layout: hybrid
+  sidePanel: true
+  showGpuStats: true
+  showCost: true
+  metrics:
+    enabled: true
+    compactMode: false
+    showPromptTokens: true
+    showTTFT: true
+    showInStatusBar: true
+  reasoning:
+    enabled: true
+    maxVisibleLines: 8
+    autoCollapseOnComplete: true
+```
+
+```bash
+# Environment variables
+export NO_COLOR=1  # Disable colored output
+
+# CLI flags
+ollm --no-color  # Disable colored output
+```
+
+### Status Settings
+
+Controls GPU and system status monitoring.
+
+| Setting                    | Type   | Default | Description                                         |
+| -------------------------- | ------ | ------- | --------------------------------------------------- |
+| `status.pollInterval`      | number | `5000`  | Status polling interval in milliseconds (min: 1000) |
+| `status.highTempThreshold` | number | `80`    | Temperature threshold for warnings in °C (min: 0)   |
+| `status.lowVramThreshold`  | number | `512`   | Low VRAM warning threshold in MB (min: 0)           |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+status:
+  pollInterval: 5000
+  highTempThreshold: 80
+  lowVramThreshold: 512
+```
+
+```bash
+# Environment variables
+# (Status settings are configured via config file)
+
+# CLI flags
+# (Status settings are not available as CLI flags)
+```
+
+### Review Settings
+
+Controls diff review behavior for file modifications.
+
+| Setting                  | Type    | Default | Description                                                   |
+| ------------------------ | ------- | ------- | ------------------------------------------------------------- |
+| `review.enabled`         | boolean | `true`  | Enable diff review for file changes                           |
+| `review.inlineThreshold` | number  | `5`     | Max lines to show inline (larger diffs open in pager, min: 0) |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+review:
+  enabled: true
+  inlineThreshold: 5
+```
+
+```bash
+# Environment variables
+# (Review settings are configured via config file)
+
+# CLI flags
+ollm --review-diffs    # Enable diff review
+ollm --no-review       # Disable diff review
+```
+
+### Session Settings
+
+Controls session recording and auto-save behavior.
+
+| Setting                | Type    | Default | Description                                    |
+| ---------------------- | ------- | ------- | ---------------------------------------------- |
+| `session.autoSave`     | boolean | `true`  | Automatically save session state               |
+| `session.saveInterval` | number  | `60000` | Auto-save interval in milliseconds (min: 1000) |
+
+**Configuration Methods:**
+
+```yaml
+# Config file
+session:
+  autoSave: true
+  saveInterval: 60000
+```
+
+```bash
+# Environment variables
+# (Session settings are configured via config file)
+
+# CLI flags
+ollm --session <id>  # Resume a specific session
+```
+
+## Environment Variables
+
+OLLM CLI supports the following environment variables for configuration:
+
+### Provider Configuration
+
+| Variable                    | Type         | Default                  | Description                                  |
+| --------------------------- | ------------ | ------------------------ | -------------------------------------------- |
+| `OLLAMA_HOST`               | string (URI) | `http://localhost:11434` | Ollama server endpoint URL                   |
+| `VLLM_HOST`                 | string (URI) | -                        | vLLM server endpoint URL                     |
+| `VLLM_API_KEY`              | string       | -                        | API key for vLLM authentication              |
+| `OPENAI_COMPATIBLE_HOST`    | string (URI) | -                        | OpenAI-compatible server endpoint URL        |
+| `OPENAI_COMPATIBLE_API_KEY` | string       | -                        | API key for OpenAI-compatible authentication |
+
+### Model Configuration
+
+| Variable             | Type   | Default       | Description               |
+| -------------------- | ------ | ------------- | ------------------------- |
+| `OLLM_DEFAULT_MODEL` | string | `llama3.2:3b` | Default model name to use |
+
+### Logging and Debugging
+
+| Variable         | Type   | Default | Description                                          |
+| ---------------- | ------ | ------- | ---------------------------------------------------- |
+| `OLLM_LOG_LEVEL` | string | `info`  | Logging verbosity (`debug`, `info`, `warn`, `error`) |
+
+### System Paths
+
+| Variable           | Type   | Default          | Description                                           |
+| ------------------ | ------ | ---------------- | ----------------------------------------------------- |
+| `OLLM_CONFIG_PATH` | string | -                | Custom config file path (overrides default locations) |
+| `OLLM_INDEX_PATH`  | string | -                | Custom codebase index location                        |
+| `XDG_CONFIG_HOME`  | string | `~/.config`      | Linux config directory (affects `~/.ollm` location)   |
+| `XDG_DATA_HOME`    | string | `~/.local/share` | Linux data directory                                  |
+| `XDG_CACHE_HOME`   | string | `~/.cache`       | Linux cache directory                                 |
+
+### Feature Flags
+
+| Variable                 | Type    | Default | Description                                  |
+| ------------------------ | ------- | ------- | -------------------------------------------- |
+| `OLLM_DISABLE_INDEXING`  | boolean | `false` | Disable automatic codebase indexing          |
+| `OLLM_DISABLE_EXECUTION` | boolean | `false` | Disable code execution sandbox               |
+| `NO_COLOR`               | boolean | `false` | Disable colored output (standard convention) |
+
+### GPU Monitoring
+
+| Variable          | Type   | Default | Description                          |
+| ----------------- | ------ | ------- | ------------------------------------ |
+| `NVIDIA_SMI_PATH` | string | -       | Custom path to nvidia-smi executable |
+
+### Usage Examples
+
+```bash
+# Set Ollama host
+export OLLAMA_HOST=http://192.168.1.100:11434
+
+# Enable debug logging
+export OLLM_LOG_LEVEL=debug
+
+# Use custom config file
+export OLLM_CONFIG_PATH=/path/to/custom/config.yaml
+
+# Disable colored output
+export NO_COLOR=1
+
+# Run with environment variables
+OLLAMA_HOST=http://localhost:11434 OLLM_LOG_LEVEL=debug ollm
+```
+
+## Configuration Examples
+
+### Minimal Configuration
+
+This is the simplest configuration that just changes the default model:
+
+```yaml
+# ~/.ollm/config.yaml or .ollm/config.yaml
+model:
+  default: llama3.1:8b
+```
+
+### Basic Configuration
+
+A typical configuration for everyday use:
+
+```yaml
+# ~/.ollm/config.yaml
+provider:
+  default: ollama
+  ollama:
+    host: http://localhost:11434
+    timeout: 30000
+
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+  maxTokens: 4096
+
+ui:
+  layout: hybrid
+  sidePanel: true
+  showGpuStats: true
+
+session:
+  autoSave: true
+  saveInterval: 60000
+```
+
+### Advanced Configuration
+
+A comprehensive configuration with all major settings:
+
+```yaml
+# ~/.ollm/config.yaml - Advanced configuration example
+
+# Provider configuration
+provider:
+  default: ollama
+  ollama:
+    host: http://localhost:11434
+    timeout: 30000
+  vllm:
+    host: http://localhost:8000
+    apiKey: your-api-key-here
+  openaiCompatible:
+    host: http://localhost:1234
+    apiKey: your-api-key-here
+
+# Model configuration
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+  maxTokens: 4096
+
+# Context management (optional - uses smart defaults if omitted)
+context:
+  targetSize: 8192
+  minSize: 2048
+  maxSize: 32768
+  autoSize: true
+  vramBuffer: 1073741824 # 1GB in bytes
+  compressionEnabled: true
+  compressionThreshold: 0.8
+  snapshotsEnabled: true
+  maxSnapshots: 5
+
+# UI configuration
+ui:
+  layout: hybrid
+  sidePanel: true
+  showGpuStats: true
+  showCost: true
+  metrics:
+    enabled: true
+    compactMode: false
+    showPromptTokens: true
+    showTTFT: true
+    showInStatusBar: true
+  reasoning:
+    enabled: true
+    maxVisibleLines: 8
+    autoCollapseOnComplete: true
+
+# Status monitoring
+status:
+  pollInterval: 5000
+  highTempThreshold: 80
+  lowVramThreshold: 512
+
+# Diff review
+review:
+  enabled: true
+  inlineThreshold: 5
+
+# Session management
+session:
+  autoSave: true
+  saveInterval: 60000
+```
+
+### Project-Specific Configuration
+
+Example workspace configuration for a coding project:
+
+```yaml
+# .ollm/config.yaml - Project-specific configuration
+
+# Use a code-specialized model for this project
+model:
+  default: codellama:13b
+  temperature: 0.3 # Lower temperature for more deterministic code generation
+  maxTokens: 8192
+
+# Enable diff review for code changes
+review:
+  enabled: true
+  inlineThreshold: 10 # Show larger diffs inline for code review
+
+# Optimize UI for coding
+ui:
+  layout: simple
+  sidePanel: false
+  showGpuStats: false
+  metrics:
+    enabled: true
+    compactMode: true
+```
+
+### Remote Server Configuration
+
+Example configuration for connecting to a remote Ollama server:
+
+```yaml
+# ~/.ollm/config.yaml - Remote server configuration
+
+provider:
+  default: ollama
+  ollama:
+    host: http://192.168.1.100:11434
+    timeout: 60000 # Longer timeout for remote connections
+
+model:
+  default: llama3.1:70b # Use larger model on powerful remote server
+  temperature: 0.7
+  maxTokens: 8192
+```
+
+### Multi-Provider Configuration
+
+Example configuration with multiple providers configured:
+
+```yaml
+# ~/.ollm/config.yaml - Multi-provider setup
+
+provider:
+  default: ollama # Default to local Ollama
+  ollama:
+    host: http://localhost:11434
+    timeout: 30000
+  vllm:
+    host: http://gpu-server:8000
+    apiKey: vllm-secret-key
+  openaiCompatible:
+    host: http://localhost:1234
+    # No API key needed for local LM Studio
+
+model:
+  default: llama3.2:3b
+  temperature: 0.7
+  maxTokens: 4096
+
+# Switch providers at runtime:
+# ollm --provider vllm --model mistral-7b
+# ollm --provider openai-compatible --model local-model
+```
+
+### Performance-Optimized Configuration
+
+Example configuration optimized for performance on limited hardware:
+
+```yaml
+# ~/.ollm/config.yaml - Performance-optimized
+
+model:
+  default: llama3.2:3b # Smaller model for faster inference
+  temperature: 0.7
+  maxTokens: 2048 # Limit output length
+
+context:
+  targetSize: 4096 # Smaller context window
+  autoSize: true
+  compressionEnabled: true
+  compressionThreshold: 0.7 # Compress earlier
+
+ui:
+  layout: simple
+  sidePanel: false
+  showGpuStats: true
+  metrics:
+    enabled: true
+    compactMode: true
+
+status:
+  pollInterval: 10000 # Poll less frequently
+```
+
+## Validation
+
+OLLM CLI validates your configuration on startup. If there are errors, you'll see detailed messages indicating:
+
+- Which file contains the error
+- The line and column number (for YAML syntax errors)
+- What's wrong and how to fix it
+- Suggestions for common mistakes
+
+Example validation error:
+
+```
+Configuration Error: Expected number, got string
+  File: /home/user/.ollm/config.yaml
+  Line: 5, Column: 15
+
+  5 | temperature: "high"
+                   ^
+
+Tip - Check for missing or unmatched quotes
+```
+
+## Tips
+
+1. **Start Simple**: Begin with a minimal configuration and add settings as needed
+2. **Use Workspace Configs**: Keep project-specific settings in `.ollm/config.yaml`
+3. **Environment Variables**: Use environment variables for temporary overrides
+4. **CLI Flags**: Use CLI flags for one-off changes without modifying config files
+5. **Validation**: Run `ollm --help` to verify your configuration loads correctly
+6. **Comments**: YAML supports comments with `#` - use them to document your settings
+
+---
+
+## Planned Future Configuration Options
+
+The following configuration options are **planned for future development** and are not yet available in the current version. These will be added as new features are implemented according to the roadmap (../DevelopmentRoadmap/Roadmap.md).
+
+### 🔮 Kraken Integration (Planned)
+
+External LLM provider configuration for accessing cloud models:
+
+```yaml
+# PLANNED - Not yet available
+# kraken:
+# enabled: true
+# confirmBeforeRelease: true
+# autoEscalation:
+# enabled: false
+# triggers:
+# - contextOverflow
+# - localModelError
+# providers:
+# geminiCli:
+# enabled: true
+# executable: gemini
+# timeout: 120000
+# defaultModel: gemini-pro
+# claudeCode:
+# enabled: true
+# executable: claude-code
+# timeout: 120000
+# openai:
+# enabled: true
+# apiKey: ${OPENAI_API_KEY}
+# model: gpt-4
+# maxTokens: 8192
+# anthropic:
+# enabled: true
+# apiKey: ${ANTHROPIC_API_KEY}
+# model: claude-3-opus
+# costTracking:
+# enabled: true
+# sessionBudget: 5.00  # USD
+# warnThreshold: 1.00
+```
+
+**Status:** Planned for Stage 10  
+**Documentation:** Kraken Integration Spec (.kiro/specs/stage-10-kraken-integration-future-dev/)
+
+### 🔮 Developer Productivity Tools (Planned)
+
+Git integration and diff review configuration:
+
+```yaml
+# PLANNED - Not yet available
+# git:
+# enabled: true
+# includeInSystemPrompt: true
+# autoCommit:
+# enabled: false
+# messageStyle: semantic  # semantic, descriptive, conventional
+# groupChanges: true
+
+# mentions:
+# enabled: true
+# maxFilesPerGlob: 50
+# maxTokensPerMention: 4096
+# warnOnLargeContext: true
+
+# diffReview:
+# enabled: true
+# autoApprove:
+# readOperations: true
+# smallChanges: true
+# smallChangeThreshold: 10  # lines
+# showFullContext: true
+# contextLines: 3
+```
+
+**Status:** Planned for Stage 11  
+**Documentation:** Developer Productivity Spec (.kiro/specs/stage-11-developer-productivity-future-dev/)
+
+### 🔮 Cross-Platform Support (Planned)
+
+Platform-specific configuration options:
+
+```yaml
+# PLANNED - Not yet available
+# platform:
+# shell:
+# windows: cmd.exe
+# unix: /bin/sh
+# python:
+# windows: python
+# unix: python3
+# paths:
+# normalizeDisplay: true  # Show forward slashes in UI
+# terminal:
+# forceUnicode: false
+# forceColor: false
+```
+
+**Status:** Planned for Stage 12  
+**Documentation:** Cross-Platform Spec (.kiro/specs/stage-12-cross-platform-future-dev/)
+
+### 🔮 File Upload System (Planned)
+
+File upload and storage configuration:
+
+```yaml
+# PLANNED - Not yet available
+# uploads:
+# enabled: true
+# methods:
+# slashCommand: true
+# clipboard: true
+# dragDrop: true
+# mentions: true
+# storage:
+# perFileLimit: 10485760  # 10MB in bytes
+# perSessionLimit: 104857600  # 100MB in bytes
+# retentionDays: 7
+# images:
+# maxWidth: 2048
+# maxHeight: 2048
+# quality: 85
+# allowedTypes:
+# - image/*
+# - text/*
+# - application/json
+```
+
+**Status:** Planned for Stage 14  
+**Documentation:** File Upload Spec (.kiro/specs/stage-14-file-upload-future-dev/)
+
+### 🔮 Intelligence Layer (Planned)
+
+Advanced AI capabilities configuration:
+
+```yaml
+# PLANNED - Not yet available
+# intelligence:
+# codebaseIndex:
+# enabled: true
+# autoIndex: true
+# extensions:
+# - .ts
+# - .js
+# - .py
+# - .java
+# - .go
+# excludePatterns:
+# - node_modules/**
+# - dist/**
+# - .git/**
+# maxFileSize: 1048576  # 1MB
+
+# structuredOutput:
+# enabled: true
+# maxRetries: 3
+# useGuidedDecoding: true
+
+# sandbox:
+# enabled: true
+# timeout: 30000
+# allowedLanguages:
+# - javascript
+# - python
+# - bash
+# restrictions:
+# networkAccess: false
+# filesystemAccess: false
+# memoryLimit: 536870912  # 512MB
+
+# vision:
+# enabled: true
+# defaultModel: llava:13b
+# maxImageSize: 5242880  # 5MB
+
+# costTracking:
+# enabled: true
+# monthlyBudget: 50.00  # USD
+# warnThreshold: 40.00
+```
+
+**Status:** Planned for Stage 15  
+**Documentation:** Intelligence Layer Spec (.kiro/specs/stage-15-intelligence-layer-future-dev/)
+
+---
+
+### Notes on Planned Features
+
+- **Not Yet Available**: These configuration options are documented for planning purposes and will be implemented in future releases
+- **Subject to Change**: Exact configuration structure may change during implementation based on technical requirements and user feedback
+- **Roadmap**: See the full roadmap (../DevelopmentRoadmap/Roadmap.md) for timeline and priority information
+- **Contributions Welcome**: If you're interested in implementing any of these features, please review the detailed specifications and open a discussion
+
+---
+
+**Last Updated:** January 15, 2026