Data Flows
This page visualizes how data moves through CLEO during key operations.
Task Creation Flow
User Input
User runs cleo add "Task title" --priority high
Input Validation
CLI validates required fields and format
ID Generation
System generates unique task ID (T###)
Atomic Write
Task written using atomic pattern
Audit Log
Operation logged to todo-log.jsonl
Atomic Write Pattern
Every write operation in CLEO follows this pattern:
This pattern ensures no partial writes - the operation either fully succeeds or the original file is preserved.
Validation Layers
Schema
Semantic
Temporal
Content
JSON Schema validation ensures correct data types, required fields, and enum values.
Cross-file checks for ID uniqueness, valid references, and cycle detection.
Timestamp validation prevents future dates and ensures logical ordering.
Content validation ensures meaningful data (title != description).
Session State Machine
Sessions transition through defined states:
Task Lifecycle
Backup System
