Skip to main content

Traycer AI

Traycer AI is an AI coding assistant that operates in two distinct modes: Phase Mode and Plan Mode.

Operational Modes

Phase Mode

Execution and implementation mode
  • Focused on carrying out planned tasks
  • Has access to phase-specific tools
  • Executes concrete actions
  • Implements solutions

Plan Mode

Planning and strategy mode
  • Focused on planning and organizing work
  • Has access to plan-specific tools
  • Creates roadmaps and strategies
  • Organizes tasks and dependencies

Architecture

Tools Structure

Traycer has separate tool sets for each mode:
  • phase_mode_tools.json: Tools available during phase/execution mode
  • plan_mode_tools.json: Tools available during planning mode

Prompts Structure

Each mode has specialized prompts:
  • phase_mode_prompts.txt: Instructions for execution mode
  • plan_mode_prompts: Instructions for planning mode

Key Characteristics

Mode Separation

  • Clear separation between planning and execution
  • Different tools and capabilities in each mode
  • Specialized instructions for each mode

Workflow

  1. Plan Mode: Analyze requirements, create strategy, organize tasks
  2. Phase Mode: Execute plan, implement solutions, complete tasks

Use Cases

Planning Mode (Plan Mode)

  • Analyze project requirements
  • Break down complex tasks
  • Create implementation roadmaps
  • Organize dependencies
  • Define milestones

Execution Mode (Phase Mode)

  • Implement planned features
  • Execute specific tasks
  • Write and modify code
  • Run tests and validations
  • Complete defined work items

Benefits

Structured Approach

  • Clear separation between thinking and doing
  • Organized workflow
  • Better task management

Specialized Tools

  • Right tools for right mode
  • Optimized for specific activities
  • Improved efficiency

Clear Context

  • Users know which mode is active
  • Appropriate expectations for each mode
  • Better collaboration

Build docs developers (and LLMs) love