Back to index
Configuration Reference
This document details all configuration parameters available in the agent-sync.yml file, their types, requirements, and descriptions.
Root Configuration
| Setting |
Type |
Required |
Description |
configVersion |
String |
Yes |
Schema version of the configuration file (e.g., “1.0”) |
projects |
Map |
No |
Map of named project configurations where keys are project identifiers (arbitrary names) and values are project configuration objects. *Typically used in standard format, omitted in simplified format. |
user |
Object |
No |
Global user-level configuration. *Typically used in standard format. |
outputDirs |
String Array |
No* |
Output directories where generated files will be placed. *Only used in simplified format. |
tasks |
Task Array |
No* |
List of generation tasks. *Only used in simplified format. |
Project Configuration
Each project can be configured with the following settings:
| Setting |
Type |
Required |
Description |
outputDirs |
String Array |
Yes |
Output directories where generated files will be placed. Supports tilde (~) expansion for home directory. Multiple directories can be specified to support scenarios like git worktree |
tasks |
Task Array |
Yes |
List of generation tasks for this project |
UserConfig Settings
User-level configuration:
| Setting |
Type |
Required |
Description |
tasks |
Task Array |
Yes |
List of generation tasks for user scope |
home |
String |
No |
Optional override for the user’s home directory, primarily for debugging purposes |
Task Configuration
Options for each task:
| Setting |
Type |
Required |
Description |
name |
String |
No |
Optional identifier for the task. If not provided, a default name is automatically generated: for project tasks, “{project-name}-{type}” (e.g., “my-project-memory”); for user tasks, “user-{type}” (e.g., “user-command”) |
type |
String |
Yes |
Type of task, one of “command”, “memory”, or “mode” |
inputs |
String Array |
Yes |
File or directory paths relative to config directory. Supports glob patterns with exclusions |
outputs |
Output Array |
Yes |
Defines the output agents and their paths |
Output Configuration
Options for each output:
| Setting |
Type |
Required |
Description |
agent |
String |
Yes |
Target AI agent (e.g., “roo”, “claude”, “cline”, “copilot”) |
outputPath |
String |
No |
Optional custom output path. If not specified, the agent’s default path is used. The path format determines concatenation behavior: paths ending with “/” are treated as directories (non-concatenated, per-file outputs), while paths without a trailing “/” are treated as files (concatenated/aggregated into a single file). Applies to all task types: memory, command, and mode. For type: mode specifically: directory outputs (e.g., Claude Code subagents/modes) generate per-mode files, while single file outputs (e.g., Roo modes) aggregate all modes into one YAML file by default. |
Navigation
Back to index