Memory Os Ai
by romainsantoli-webcommunityv3.0.1
1
Memory OS AI is an adaptive memory system for AI agents that transforms local documents into a semantic memory queryable by any MCP-compatible client through the Model Context Protocol.
From the registry: Adaptive memory for AI agents — FAISS search, chat extraction, cross-project linking
Install & configure
Prerequisites
- requiredPython 3.10+Python runtime
- requiredClone and installClone repo and install via pip
$ git clone https://github.com/romainsantoli-web/Memory-os-ai.git && cd Memory-os-ai && pip install -e '.[dev,audio]'
Install in your client
Paste into the AI you're chatting with — it will figure out the rest
Please install the `memory-os-ai` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Python 3.10+** — Python runtime
- **Clone and install** — Clone repo and install via pip Run: `git clone https://github.com/romainsantoli-web/Memory-os-ai.git && cd Memory-os-ai && pip install -e '.[dev,audio]'`
Canonical MCP server config (stdio transport):
- command: `memory-os-ai`
- args: []
- optional environment variables:
- `MEMORY_CACHE_DIR`: Cache/FAISS index directory (example: `~/.memory-os-ai`)
- `MEMORY_MODEL`: SentenceTransformer model name (example: `all-MiniLM-L6-v2`)
- `MEMORY_API_KEY`: Optional API key for SSE/HTTP auth
- `MEMORY_CLOUD_PROVIDER`: Cloud provider name (example: `icloud`)
- `MEMORY_CLOUD_CONFIG`: JSON credentials
- `MEMORY_DISK_THRESHOLD`: Bytes free before cloud overflow (example: `524288000`)
Note: 21 MCP tools. Supports stdio/SSE/HTTP transports.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.Note: 21 MCP tools. Supports stdio/SSE/HTTP transports.
Tools (21)
memory_ingest
Index a folder of documents into FAISS
memory_search
Semantic search across all indexed content
memory_search_occurrences
Count keyword occurrences across documents
memory_get_context
Get relevant context for the current task
memory_list_documents
List all indexed documents with stats
memory_transcribe
Transcribe audio files (Whisper)
memory_status
Engine status (index size, model, device)
memory_compact
Compact/deduplicate the FAISS index
memory_chat_sync
Sync messages from configured chat sources
memory_chat_source_add
Add a chat source (Claude, ChatGPT, etc.)
memory_chat_source_remove
Remove a chat source
memory_chat_status
Status of all chat sources
memory_chat_auto_detect
Auto-detect chat workspaces on disk
memory_session_brief
Full memory briefing for session start
Environment variables
MEMORY_API_KEYOptional API key for SSE/HTTP authMEMORY_CLOUD_PROVIDERrequiredCloud provider name (see table above)MEMORY_CLOUD_CONFIGrequiredJSON credentials or path to JSON filePrerequisites
- • Python 3.10+
Packages
memory-os-ai
pypiv3.0.1transport: stdio
Details
- Language
- Python
- License
- GPL-3.0
- Released
- Mar 3, 2026
- Last commit
- Mar 3, 2026
- Registry name
- io.github.romainsantoli-web/memory-os-ai