Add Action Memory System

[trmxvibs]

Objective

Create an in-memory action history system for OpenOperator.

The system should record every action executed by the agent so that future components can inspect previous behavior.

This is the first step toward:

• Execution history
• Replay functionality
• Audit logs
• Self-correction
• Autonomous workflows

Requirements

Create:

src/openoperator/core/action_memory.py

Models

ActionRecord

Fields:

• step: int
• action: str
• status: str
• timestamp: str

Memory Class

ActionMemory

Methods:

• add_record(record)
• get_records()
• clear()

Requirements:

• In-memory only
• No database
• No file storage
• Python 3.11+
• Pydantic v2 compatible
• Fully typed
• Contributor-friendly implementation

Acceptance Criteria

The following should work:

memory.add_record(...)
memory.get_records()
memory.clear()

Records must remain available during runtime until cleared.

[trmxvibs]

Implemented ActionMemory system.

Features:

• ActionRecord model
• ActionMemory storage
• add_record()
• get_records()
• clear()

Verified locally using action_memory_demo.py.

Closing issue.