claude-configs/skills/memorygraph/SKILL.md

name	description
memorygraph	Graph-based memory system for AI-assisted pattern recall, solution matching, and knowledge relationships. USE WHEN storing learnings, recalling past solutions, linking problems to fixes, tracking decisions, or building knowledge connections across sessions.

MemoryGraph - AI Memory Mind Map

Purpose

MemoryGraph provides persistent, graph-based memory for pattern recognition and solution recall. While NoteDiscovery handles long-form knowledge articles, MemoryGraph creates a web of interconnected memories optimized for AI retrieval.

Use MemoryGraph for:

• Quick learnings and solutions (not full articles)
• Problem/solution pairs with relationships
• Decisions and their rationale
• Patterns that should be recalled automatically
• Cross-session knowledge accumulation

Use NoteDiscovery for:

• Detailed documentation and guides
• Reference material for humans
• Structured knowledge articles
• Searchable documentation

When to Activate This Skill

Explicit triggers:

• "Remember this for later"
• "Store this solution"
• "What did we learn about X?"
• "Find solutions for this problem"
• "Link this to that issue"
• "What causes this error?"
• "Show me related memories"

Automatic triggers (per CLAUDE.md Memory Protocol):

• After fixing a bug
• After a git commit
• After making an architecture decision
• After discovering a reusable pattern
• After a successful configuration
• After troubleshooting sessions

Quick Reference

CLI Commands

# Store a memory
python client.py store --type solution --title "Fixed X" --content "Details..." --tags "tag1,tag2" --importance 0.8

# Recall memories (natural language search)
python client.py recall "timeout error redis"

# Get specific memory
python client.py get <memory_id>

# Create relationship
python client.py relate <from_id> <to_id> SOLVES --context "How it solves"

# Search with filters
python client.py search --query "authentication" --types "solution,fix" --tags "python"

# Update memory
python client.py update <memory_id> --title "New title" --importance 0.9

# Delete memory
python client.py delete <memory_id> --force

# Get related memories
python client.py related <memory_id> --types "SOLVES,CAUSES" --depth 2

# Statistics
python client.py stats

# Recent activity
python client.py recent --days 7

Memory Types

Type	When to Use
solution	Fix or resolution to a problem
problem	Issue or challenge encountered
error	Specific error message or condition
fix	Code-level patch or correction
decision	Architectural or design choice
code_pattern	Reusable code pattern
configuration	Config that worked
workflow	Process or sequence
general	Catch-all for other learnings

Relationship Types

Solution relationships:

• SOLVES - Solution fixes a problem
• FIXES - Fix addresses an error
• ADDRESSES - General resolution
• IMPROVES - Enhancement over previous

Causal relationships:

• CAUSES - A causes B
• TRIGGERS - A triggers B
• PREVENTS - A prevents B

Dependency relationships:

• REQUIRES - A needs B
• DEPENDS_ON - A depends on B
• BUILDS_ON - A extends B

Association:

• RELATED_TO - General association

Importance Scale

Score	Meaning
0.8-1.0	Critical - affects multiple projects, prevents major issues
0.5-0.7	Standard - useful pattern or solution
0.3-0.4	Minor - nice-to-know, edge cases

Workflow Patterns

1. Store a Bug Fix

# Store the problem
python client.py store --type problem \
  --title "Redis connection timeouts in production" \
  --content "Connections timing out after 30s idle..." \
  --tags "redis,timeout,production" \
  --importance 0.7

# Store the solution
python client.py store --type solution \
  --title "Configure Redis keepalive and timeout" \
  --content "Added socket_keepalive=True and socket_timeout=300..." \
  --tags "redis,configuration" \
  --importance 0.8

# Link them
python client.py relate <solution_id> <problem_id> SOLVES \
  --context "Keepalive prevents idle disconnections"

2. Recall Before Starting Work

# Check what we know about a topic
python client.py recall "authentication oauth"

# Get recent activity
python client.py recent --days 7

# Find solutions for similar problems
python client.py search --types "solution" --tags "python,api"

3. Document a Decision

python client.py store --type decision \
  --title "Chose PostgreSQL over MongoDB" \
  --content "Rationale: Need ACID transactions, complex queries..." \
  --tags "database,architecture,decision" \
  --importance 0.9

4. Track Error Patterns

# Store error
python client.py store --type error \
  --title "ModuleNotFoundError: No module named 'xyz'" \
  --content "Occurs when running in venv without dependencies..." \
  --tags "python,venv,import" \
  --importance 0.5

# Store fix
python client.py store --type fix \
  --title "Reinstall dependencies in active venv" \
  --content "pip install -e . or pip install -r requirements.txt" \
  --tags "python,venv" \
  --importance 0.5

# Link
python client.py relate <fix_id> <error_id> FIXES

Python API Usage

from client import MemoryGraphClient

client = MemoryGraphClient()

# Store
memory_id = client.store_memory(
    type="solution",
    title="Redis timeout fix",
    content="Configure keepalive...",
    tags=["redis", "timeout"],
    importance=0.8
)

# Recall
memories = client.recall_memories("redis timeout", limit=5)

# Get with relationships
memory = client.get_memory(memory_id, include_relationships=True)

# Create relationship
rel_id = client.create_relationship(
    from_memory_id=solution_id,
    to_memory_id=problem_id,
    relationship_type="SOLVES",
    context="Fixes the timeout issue"
)

# Find related
related = client.get_related_memories(
    memory_id,
    relationship_types=["SOLVES", "CAUSES"],
    max_depth=2
)

# Stats
stats = client.get_memory_statistics()

Integration with NoteDiscovery

These tools complement each other:

Scenario	Use
Quick bug fix	MemoryGraph (store_memory type=fix)
Troubleshooting guide	NoteDiscovery (detailed article)
Decision rationale	MemoryGraph (type=decision)
Architecture documentation	NoteDiscovery
Pattern discovered	MemoryGraph (type=code_pattern)
Best practices guide	NoteDiscovery
Error/solution pair	MemoryGraph (with SOLVES relationship)

Workflow: After creating memories in MemoryGraph, consider if a fuller NoteDiscovery article would help. Link them conceptually via tags.

Database Location

• SQLite Database: ~/.memorygraph/memory.db
• Client: ~/.claude/skills/memorygraph/client.py
• Schema: ~/.claude/skills/memorygraph/SCHEMA.md

Proactive Usage

This skill should be used proactively when:

1. Bug fixed - Store problem + solution + SOLVES relationship
2. Git commit made - Store summary of what was fixed/added
3. Architecture decision - Store choice + rationale
4. Pattern discovered - Store for future recall
5. Configuration worked - Store what worked and why
6. Troubleshooting complete - Store what was tried, what worked
7. Session ending - Prompt: "Should I store today's learnings?"

---

Location: ~/.claude/skills/memorygraph/ Version: 1.0.0 Created: 2025-12-07