SuperLocalMemory implements the Model Context Protocol natively. Install once, integrations auto-configure across supported AI tools.
Auto-configured on install. No manual JSON editing for most tools.
+ any custom MCP client via Python SDK or HTTP transport
100% local — zero data ever leaves your machine
remember() Save memories with tags and project context
recall() Semantic search across all memories
search() OpenAI MCP spec search (ChatGPT compatible)
fetch() Retrieve a specific memory by ID
list_recent() Get recent memories with metadata
get_status() System health and statistics
build_graph() Build knowledge graph from memories
switch_profile() Switch between user profiles
backup_status() Auto-backup health check
memory_used() Implicit feedback for adaptive learning
get_learned_patterns() View what the system learned about you
correct_pattern() Override a learned preference
session_init() Auto-recall project context at session start (V3.1)
observe() Auto-capture decisions, bugs, preferences from conversation (V3.1)
report_feedback() Explicit feedback to train adaptive ranking (V3.1)
slm://context Active session context — auto-injected on connect (V3.1)
memory://recent/{limit} Recent memories stream
memory://stats System statistics
memory://graph/clusters Knowledge graph clusters
memory://patterns/identity Your coding identity
memory://learning/status Learning system state
memory://engagement Engagement metrics
+ 2 MCP Prompts: coding_identity_prompt and project_context_prompt
Auto-detects and configures all supported IDEs. Requires Node.js 18+.