Goal Plan

Features

• Goal-Oriented Action Planning (GOAP)
• Precondition analysis and effect tracking
• Cost estimation for actions
• Adaptive replanning on failure or change
• Structured task creation and execution

Use Cases

• Breaking down large software development tasks into executable steps.
• Automating complex research or data analysis projects.
• Creating adaptive strategies for long-term project management.
• Developing AI agents that can self-correct and replan when encountering obstacles.

Non-Goals

• Directly executing arbitrary shell commands without a planning context.
• Managing individual tasks outside of a larger GOAP-defined plan.
• Providing a generic task runner without the planning and replanning logic.

Workflow

1. Define goal state and current state.
2. Identify the gap between current and goal states.
3. Inventory available actions with preconditions, effects, and costs.
4. Generate an optimal action sequence plan using A* search.
5. Record the start of the planning trajectory.
6. Create tasks for each action in the plan.
7. Execute actions sequentially, verifying preconditions and effects.
8. Record each execution step.
9. Monitor for failures or unexpected results and replan if necessary.
10. Complete the planning trajectory upon success.
11. Store the successful plan in memory.

Practices

• Action Planning
• AI Orchestration
• Adaptive Systems

Prerequisites

• Access to Claude Code MCP tools (claude-flow)
• A defined goal description as input