> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/skydiscover-ai/skydiscover/llms.txt
> Use this file to discover all available pages before exploring further.

# Configuration Overview

> Learn how to configure SkyDiscover using YAML files and understand the configuration hierarchy

## Introduction

SkyDiscover uses YAML configuration files to control all aspects of the evolutionary search process. Configuration files define LLM settings, search algorithms, evaluation parameters, prompts, and more.

## Configuration Structure

A SkyDiscover configuration file consists of several top-level sections:

```yaml theme={null}
# General settings
max_iterations: 100
checkpoint_interval: 10
log_level: "INFO"

# Component configurations
llm:          # LLM model settings
prompt:       # Prompt generation
search:       # Search algorithm and database
evaluator:    # Program evaluation
agentic:      # Agentic generation (optional)
monitor:      # Live dashboard (optional)
```

## Loading Configuration

### From File

Load a configuration file when running SkyDiscover:

```bash theme={null}
skydiscover-run initial_program.py evaluator.py -c configs/adaevolve.yaml
```

### Programmatic Loading

```python theme={null}
from skydiscover.config import load_config

# Load from YAML file
config = load_config("configs/adaevolve.yaml")

# Use defaults
config = load_config()
```

## Configuration Hierarchy

SkyDiscover resolves configuration values in the following order (later sources override earlier ones):

<Steps>
  <Step title="Default Values">
    Built-in defaults from dataclass definitions in `skydiscover/config.py:522-561`
  </Step>

  <Step title="YAML File">
    Values specified in your configuration file override defaults
  </Step>

  <Step title="Environment Variables">
    Environment variables like `OPENAI_API_KEY`, `OPENAI_API_BASE` override file settings
  </Step>

  <Step title="CLI Arguments">
    Command-line flags like `--model`, `--search` override all previous settings
  </Step>
</Steps>

### Environment Variable Expansion

Use `${VAR}` syntax to reference environment variables in YAML:

```yaml theme={null}
llm:
  api_key: ${OPENAI_API_KEY}
  models:
    - name: "gpt-5"
```

## General Settings

<ParamField path="max_iterations" type="int" default="100">
  Maximum number of evolutionary iterations to run
</ParamField>

<ParamField path="checkpoint_interval" type="int" default="10">
  Save checkpoint every N iterations
</ParamField>

<ParamField path="log_level" type="str" default="INFO">
  Logging verbosity: `DEBUG`, `INFO`, `WARNING`, `ERROR`
</ParamField>

<ParamField path="log_dir" type="str" default="None">
  Directory for log files. If None, logs to console only
</ParamField>

<ParamField path="language" type="str" default="None">
  Programming language hint (e.g., `python`, `javascript`)
</ParamField>

<ParamField path="file_suffix" type="str" default=".py">
  File extension for generated programs
</ParamField>

## Generation Settings

<ParamField path="diff_based_generation" type="bool" default="true">
  Generate diffs instead of complete programs to improve LLM focus
</ParamField>

<ParamField path="max_solution_length" type="int" default="60000">
  Maximum character length for generated solutions
</ParamField>

<ParamField path="max_parallel_iterations" type="int" default="1">
  Number of iterations to run concurrently. Set to >1 for parallel execution
</ParamField>

## Human-in-the-Loop Settings

<ParamField path="human_feedback_enabled" type="bool" default="false">
  Enable human feedback integration
</ParamField>

<ParamField path="human_feedback_file" type="str" default="None">
  Path to file containing human feedback
</ParamField>

<ParamField path="human_feedback_mode" type="str" default="append">
  How to handle feedback: `append` or `replace`
</ParamField>

## Configuration Files

SkyDiscover includes several preset configurations in `configs/`:

<CardGroup cols={2}>
  <Card title="default.yaml" icon="star" href="https://github.com/yourusername/skydiscover/blob/main/configs/default.yaml">
    Basic top-k search configuration
  </Card>

  <Card title="adaevolve.yaml" icon="island" href="https://github.com/yourusername/skydiscover/blob/main/configs/adaevolve.yaml">
    Adaptive multi-island evolutionary search
  </Card>

  <Card title="openevolve_native.yaml" icon="grid" href="https://github.com/yourusername/skydiscover/blob/main/configs/openevolve_native.yaml">
    MAP-Elites quality-diversity search
  </Card>

  <Card title="llm_judge.yaml" icon="gavel" href="https://github.com/yourusername/skydiscover/blob/main/configs/llm_judge.yaml">
    LLM-as-a-judge evaluation
  </Card>
</CardGroup>

## Example: Complete Configuration

```yaml configs/default.yaml theme={null}
# General settings
max_iterations: 100
checkpoint_interval: 10
log_level: "INFO"
random_seed: 42

# LLM configuration
llm:
  models:
    - name: "gpt-5"
      weight: 1.0
  temperature: 0.7
  top_p: 0.95
  max_tokens: 32000
  timeout: 600

# Search configuration
search:
  type: "topk"
  database:
    random_seed: 42
  num_context_programs: 4

# Prompt configuration
prompt:
  system_message: "You are an expert to help find the best solution to the problem."

# Evaluator configuration
evaluator:
  timeout: 10000
  max_retries: 3
  cascade_evaluation: false

# Generation settings
diff_based_generation: true
max_solution_length: 60000
```

## Next Steps

<CardGroup cols={2}>
  <Card title="LLM Configuration" icon="brain" href="/config/llm">
    Configure models, API settings, and generation parameters
  </Card>

  <Card title="Search Configuration" icon="magnifying-glass" href="/config/search">
    Choose and configure search algorithms
  </Card>

  <Card title="Prompt Configuration" icon="message" href="/config/prompt">
    Customize system messages and prompts
  </Card>

  <Card title="Monitor Configuration" icon="chart-line" href="/config/monitor">
    Set up the live monitoring dashboard
  </Card>
</CardGroup>