memory-retrieve

# Memory Retrieve Skill

Load and apply relevant memories at session start and during conversations.

## Context Loading at Session Start

```python
from memory_lib import load_session_context

# Load all relevant context for current session
context = load_session_context("/path/to/project")

# Access different memory categories
context['global_memories']       # All global memories
context['project_memories']      # Project-specific memories
context['recent_memories']       # Recently accessed memories
context['high_confidence_memories']  # Confidence > 0.8
```

## Retrieving Relevant Memories

```python
from memory_lib import search_memories, load_session_context

# Get current project context
context = load_session_context(".")
project_hash = context.get('project_hash')

# Search for relevant memories
relevant = search_memories(
    query="user is asking about testing",
    project_hash=project_hash,
    limit=5,
    min_confidence=0.5
)

# Apply the most relevant memories
for memory in relevant:
    apply_memory(memory)
```

## Formatting Memories for Display

```python
def format_memory_for_display(memory):
    """Format a single memory for display."""
    content = memory['content']
    meta = memory['metadata']

    output = f"### {content['title']}\n"
    output += f"**Confidence:** {meta['confidence']:.0%}\n"
    output += f"**Scope:** {memory['scope']['type']}\n\n"
    output += f"{content['description']}\n\n"
    output += f"**Action:** {content['action']}\n"

    if content.get('examples'):
        output += "\n**Examples:**\n"
        for ex in content['examples']:
            output += f"- {ex}\n"

    return output

def format_memories_for_context(memories):
    """Format multiple memories for injection into context."""
    if not memories:
        return ""

    output = "## Relevant Memories\n\n"
    for memory in memories:
        output += format_memory_for_display(memory)
        output += "\n---\n\n"

    return output
```

## Applying High-Conf