32 AI Memory Tools

Comprehensive vector memory system with persistent storage, semantic search, knowledge graphs, and multi-agent coordination.

160,427+
Vector Memories
8ms
Search Latency
32
API Tools
8
Feature Categories

1. Core Memory Operations

Essential memory storage, retrieval, and management operations forming the foundation of the Memory Spine system.

memory_store
Ingest new memories via /ingest endpoint with automatic vectorization and tagging.
POST /ingest → Vector embedding → SQLite storage
memory_search
Hybrid FTS5 + semantic search with query expansion and relevance scoring.
GET /search?q={query}&limit={n} → Ranked results
memory_retrieve
Get memory by exact UUID with full content and metadata.
GET /memory/{id} → Complete memory object
memory_recent
Get last N memories ordered by creation timestamp.
GET /recent?count={n} → Chronological list
memory_delete
Permanently remove memory by ID with cascade deletion of related data.
DELETE /memory/{id} → Confirmation
memory_update
Modify existing memory content with re-vectorization and version tracking.
PUT /memory/{id} → Updated memory
memory_pin
Pin critical memories (system config, goals) for persistence across cleanup operations.
POST /pin/{key} → Pinned memory reference
memory_get_pin
Retrieve pinned memory by key for quick access to critical system data.
GET /pin/{key} → Pinned memory content

Core Memory Example

// Store a new memory
await fetch('/ingest', {
  method: 'POST',
  body: JSON.stringify({
    content: "User prefers dark mode UI",
    tags: ["preferences", "ui"],
    metadata: { priority: "high" }
  })
});

// Search memories
const results = await fetch('/search?q=dark mode&limit=5')
  .then(r => r.json());

// Pin critical config
await fetch('/pin/user-preferences', {
  method: 'POST',
  body: JSON.stringify({
    content: "System configuration for user preferences",
    key: "user-preferences"
  })
});

2. Advanced Analysis

Sophisticated memory analysis tools for patterns, clustering, summarization, and intelligent consolidation.

memory_analytics
Usage patterns, tag distribution, growth metrics, and system health analysis.
GET /analytics → Comprehensive statistics
memory_timeline
Query memories within specific time ranges with hourly precision.
GET /timeline?hours={n} → Time-filtered memories
memory_find_similar
Cosine similarity search to find related memories with confidence scores.
GET /similar/{id}?limit={n} → Similar memories
memory_cluster
Cluster memories by tags, topics, or semantic similarity for organization.
GET /cluster?group_by={field} → Clustered groups
memory_summarize
AI-powered summarization of memory collections on specific topics.
POST /summarize → Intelligent summary
memory_consolidate
Merge old/redundant memories with configurable decay threshold (0.0-1.0).
POST /consolidate?threshold={n} → Cleanup report

Advanced Analysis Example

// Get analytics dashboard
const analytics = await fetch('/analytics').then(r => r.json());
console.log(`Total memories: ${analytics.total_count}`);
console.log(`Top tags: ${analytics.tag_distribution.slice(0, 5)}`);

// Find similar memories
const similar = await fetch('/similar/uuid-123?limit=3')
  .then(r => r.json());

// Consolidate old memories (70% similarity threshold)
const consolidation = await fetch('/consolidate?threshold=0.7', {
  method: 'POST'
}).then(r => r.json());
console.log(`Merged ${consolidation.merged_count} redundant memories`);

3. Knowledge Graph

Automatically generate and query knowledge graphs from your memory collections for relationship discovery.

knowledge_graph_build
Auto-generate knowledge graph from memories with entity extraction and relationship mapping.
POST /graph/build?limit={n} → Graph structure
knowledge_graph_query
Traverse and query the knowledge graph for complex relationship queries.
GET /graph/query/{memory_id} → Graph traversal

Knowledge Graph Example

// Build knowledge graph from recent memories
const graph = await fetch('/graph/build?limit=100', {
  method: 'POST'
}).then(r => r.json());

console.log(`Generated graph with ${graph.nodes} nodes, ${graph.edges} edges`);

// Query relationships
const relationships = await fetch('/graph/query/uuid-123')
  .then(r => r.json());

relationships.connections.forEach(conn => {
  console.log(`${conn.source} → ${conn.relation} → ${conn.target}`);
});

4. Conversation Intelligence

Track, analyze, and persist multi-turn conversations with context management and long-term memory integration.

conversation_start
Begin tracked conversation with system prompt and session initialization.
POST /conversation/start → Session ID
conversation_add_turn
Add user/assistant turns to active conversation with automatic context tracking.
POST /conversation/turn → Updated context
conversation_get_context
Get conversation context window with configurable token limits.
GET /conversation/context?tokens={n} → Context
conversation_save
Persist conversation to long-term memory with tags and metadata.
POST /conversation/save → Memory ID

Conversation Intelligence Example

// Start a tracked conversation
const session = await fetch('/conversation/start', {
  method: 'POST',
  body: JSON.stringify({
    system_prompt: "You are a helpful AI assistant focused on code review."
  })
}).then(r => r.json());

// Add conversation turns
await fetch('/conversation/turn', {
  method: 'POST',
  body: JSON.stringify({
    role: "user",
    content: "Please review this JavaScript function for security issues."
  })
});

// Get context for LLM
const context = await fetch('/conversation/context?tokens=4000')
  .then(r => r.json());

// Save to long-term memory
await fetch('/conversation/save', {
  method: 'POST',
  body: JSON.stringify({
    tags: ["code-review", "security", "javascript"]
  })
});

5. Codebase Integration

Intelligent codebase analysis and context generation for AI-assisted development workflows.

codebase_context
Pre-edit context loading with relevant file analysis and dependency mapping.
GET /codebase/context?path={path}&task={task}
codebase_analyze
Deep repository analysis with complexity metrics and architecture insights.
POST /codebase/analyze → Analysis report
codebase_suggest
AI-powered improvement suggestions based on codebase patterns and best practices.
GET /codebase/suggest?focus={area} → Suggestions

Codebase Integration Example

// Get context for editing a specific file
const context = await fetch('/codebase/context?' + 
  'path=/src/auth.js&task=add OAuth integration')
  .then(r => r.json());

// Analyze entire repository
const analysis = await fetch('/codebase/analyze', {
  method: 'POST',
  body: JSON.stringify({ path: '/project/root' })
}).then(r => r.json());

console.log(`Complexity score: ${analysis.complexity}`);
console.log(`Architecture: ${analysis.architecture_patterns}`);

// Get improvement suggestions
const suggestions = await fetch('/codebase/suggest?focus=security')
  .then(r => r.json());

6. Multi-Agent Coordination

Seamless context handoff and batch operations for multi-agent AI systems and workflow automation.

agent_handoff
Transfer context between agents with recent memory inclusion and target specification.
POST /agent/handoff → Context package
memory_batch_store
Store multiple memories simultaneously for efficient bulk operations.
POST /batch/store → Batch results
memory_batch_tag
Tag multiple memories simultaneously for efficient categorization workflows.
PUT /batch/tag → Tag results

Multi-Agent Example

// Hand off context to specialized agent
const handoff = await fetch('/agent/handoff', {
  method: 'POST',
  body: JSON.stringify({
    target_agent: "code-reviewer",
    include_recent: 10
  })
}).then(r => r.json());

// Batch store memories from multiple sources
const batch_memories = [
  { content: "Code review findings", tags: ["review", "security"] },
  { content: "Performance optimizations", tags: ["perf", "optimization"] },
  { content: "Bug fixes applied", tags: ["bugs", "fixes"] }
];

await fetch('/batch/store', {
  method: 'POST',
  body: JSON.stringify({ memories: batch_memories })
});

// Batch tag related memories
await fetch('/batch/tag', {
  method: 'PUT',
  body: JSON.stringify({
    memory_ids: ["uuid-1", "uuid-2", "uuid-3"],
    add_tags: ["sprint-2", "release-candidate"]
  })
});

7. Query Power Tools

Advanced query capabilities with custom DSL syntax and optimized LLM context window generation.

memory_query_dsl
Custom DSL syntax for complex queries with boolean operators and filters.
POST /query/dsl → Query results
llm_context_window
Build optimized LLM context from relevant memories with token budget management.
GET /llm/context?q={query}&tokens={n}

Query Power Example

// Advanced DSL query
const dsl_results = await fetch('/query/dsl', {
  method: 'POST',
  body: JSON.stringify({
    dsl_query: "tags:security AND NOT tags:resolved AND created:last-week"
  })
}).then(r => r.json());

// Optimized LLM context generation
const llm_context = await fetch('/llm/context?' + 
  'q=authentication bugs&tokens=8000')
  .then(r => r.json());

console.log(`Context: ${llm_context.content}`);
console.log(`Token count: ${llm_context.token_count}`);
console.log(`Relevance score: ${llm_context.relevance}`);

8. System Operations

System monitoring, health diagnostics, and administrative tools for Memory Spine management.

memory_stats
Vector count, storage usage, and system performance statistics.
GET /stats → System metrics
memory_health_check
Comprehensive health diagnostics including database integrity and performance.
GET /health → Health report
memory_context
Build general context from memories for various AI applications.
GET /context?q={query}&tokens={n}
memory_unpin
Remove pinned memories to allow cleanup and reorganization.
DELETE /pin/{key} → Unpin confirmation

System Operations Example

// Check system statistics
const stats = await fetch('/stats').then(r => r.json());
console.log(`Total memories: ${stats.total_memories}`);
console.log(`Storage used: ${stats.storage_mb}MB`);
console.log(`Avg search time: ${stats.avg_search_ms}ms`);

// Health check
const health = await fetch('/health').then(r => r.json());
if (health.status === 'healthy') {
  console.log('All systems operational');
} else {
  console.warn('Issues detected:', health.issues);
}

// General context generation
const context = await fetch('/context?q=recent bug reports&tokens=2000')
  .then(r => r.json());

Technical Specifications

Backend Database SQLite + FTS5
Vector Count 160,427+
Search Latency 8ms average
API Protocol REST + MCP
Tag System Flexible + Batch Ops
Deduplication Similarity Detection
Decay Threshold 0.0-1.0 configurable
Context Window Up to 32K tokens

Ready to Transform Your AI Memory?

Join thousands of developers using Memory Spine for persistent AI memory and intelligent context management.

Start Building with ChaozCode Explore Full Platform See Pricing

Memory Spine is included with every ChaozCode plan

Free
5K vectors
$0/mo
POPULAR
Indie
25K vectors
$19/mo
Developer
100K vectors
$49/mo
Master
250K vectors
$99/mo
Compare all plans →