Skip to content
Agentmart
Submit a tool

Memory OS AI

Adaptive memory for AI agents — FAISS search, chat extraction, cross-project linking

Connector by romainsantoli-web · GitHub account verified by the MCP Registry

IndexedGPL-3.0Local (stdio)★ 3
Write a review

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/romainsantoli-web-memory-os-ai

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Adaptive memory for AI agents — FAISS search, chat extraction, cross-project linking

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list, install config and code. Checked Oct 25, 2026.

Tools
11 read · 2 write · 1 delete · 7 not yet classified
Secrets
None asked for
Install scripts
None found
Runs as
Python package

Tools by what they do 21

Read from the source code; the running server may list different tools.

Read 11 · Look things up without changing anything.

  • memory_chat_auto_detect
  • memory_chat_source_remove
  • memory_chat_status
  • memory_chat_sync
  • memory_cloud_status
  • memory_get_context
  • memory_list_documents
  • memory_project_list
  • memory_search
  • memory_search_occurrences
  • memory_status

Write 2 · Create, change, send or run things.

  • memory_cloud_sync
  • memory_session_brief

Delete 1 · Delete, drop, revoke or wipe things.

  • memory_project_unlink

Not yet classified 7 · We could not tell from the name and description.

  • memory_chat_save
  • memory_chat_source_add
  • memory_cloud_configure
  • memory_compact
  • memory_ingest
  • memory_project_link
  • memory_transcribe

Network hosts it talks to 4

  • github.com · found in code
  • googleapis.com · found in code
  • graph.microsoft.com · found in code
  • huggingface.co · found in code

Files it touches

  • /Users/.../OneDrive
  • /Users/me/other-project/.memory-os-ai
  • /Users/romain
  • /Users/romain/.memory-os-ai
  • /Users/romain/Memory-os-ai/.venv/bin/python
  • /Users/romain/Memory-os-ai/src
  • Your home directory (~)
  • ~/.config/Code/User/workspaceStorage
  • ~/.memory-os-ai
  • ~/.vscode-server/data/User/workspaceStorage
  • ~/Library/CloudStorage/OneDrive
  • ~/Library/CloudStorage/OneDrive-Personal
  • ~/OneDrive

Tools (21)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

  • memory_chat_auto_detect

    Automatically detect all VS Code workspace storage directories on this machine and optionally register them as chat sources. Discovers all Copilot chat histories without manual configuration.

  • memory_chat_save

    CALL THIS TOOL REGULARLY during the conversation to persist the current exchange into semantic memory. Call it: (1) after completing a significant task, (2) every ~10 exchanges, (3) when the user asks you to remember something, (4) before ending a session. Messages are indexed in FAISS AND saved to a persistent JSONL log so they survive process restarts. Include a summary for fast recall.

  • memory_chat_source_add

    Register a new chat source to extract from. Types: 'vscode' (VS Code Copilot DB), 'jsonl' (JSONL log file), 'markdown' (MD export), 'folder' (watch folder for new files).

  • memory_chat_source_remove

    Unregister a chat source by its ID.

  • memory_chat_status

    Show all registered chat sources, their sync state, and the number of messages extracted so far.

  • memory_chat_sync

    Incrementally extract new chat messages from all registered sources (VS Code Copilot, JSONL, Markdown, folder) and feed them into the semantic memory. Only new/changed content is processed.

  • memory_cloud_configure

    Configure a cloud storage backend for memory overflow when local disk is low. Supported providers: Google Drive, iCloud, Dropbox, OneDrive, Amazon S3, Azure Blob, Box, Backblaze B2. Once configured, memory data automatically syncs to cloud when disk space drops below threshold (default 500 MB).

  • memory_cloud_status

    Show local disk usage, cloud storage status, quota, and available providers. Indicates whether disk is low and whether cloud overflow is active.

  • memory_cloud_sync

    Synchronize memory data between local disk and cloud. Directions: 'push' (upload local→cloud), 'pull' (download cloud→local), 'auto' (check disk, offload if needed). Use 'push' for backup, 'pull' to restore on a new machine, 'auto' for hands-free overflow.

  • memory_compact

    Compress the in-memory FAISS index by removing near-duplicate segments and merging short fragments. Use when the memory is getting large and context retrieval is slow or noisy. Strategies: 'dedup_merge' (default) removes duplicates and merges short segments; 'top_k' keeps only the most central segments.

  • memory_get_context

    Retrieve relevant document context for a query. Designed to provide context for summarization, report generation, or Q&A — the AI model uses this to ground responses in real documents.

  • memory_ingest

    Ingest documents from a folder into the semantic memory. Supports PDF, TXT, DOCX, DOC, PNG, JPEG, PPT, PPTX, MP3, WAV, OGG, FLAC. Documents are segmented, embedded via SentenceTransformers, and indexed in FAISS.

  • memory_list_documents

    List all indexed documents with optional statistics (page count, segment count, word count).

  • memory_project_link

    Link another project's memory cache so that searches and session briefs include context from both projects. Enables cross-project knowledge sharing. Provide the absolute path to the other project's .memory-os-ai directory.

  • memory_project_list

    List all currently linked external project memories with their status.

Show 6 more tools
  • memory_project_unlink

    Remove a previously linked project by its alias.

  • memory_search

    Search indexed documents using natural language. Returns the most relevant text segments with their source file and distance score.

  • memory_search_occurrences

    Count exact keyword occurrences across indexed documents. Uses FAISS for candidate selection then regex for precise counting.

  • memory_session_brief

    CALL THIS TOOL AT THE START OF EVERY CONVERSATION. Also call it whenever you feel you have lost context or the user mentions something you should remember but don't. Generates a comprehensive session briefing from the semantic memory: project overview, recent activity, pending tasks, key context, and chat history. Optionally syncs chat sources first and accepts a focus query to prioritise specific topics.

  • memory_status

    Return the current status of the Memory OS AI engine: device, model, index size, document count.

  • memory_transcribe

    Transcribe an audio file (MP3, WAV, OGG, FLAC) to text using OpenAI Whisper. The transcript can then be ingested or analyzed by Copilot.

Compare and alternatives

    Featured in

    Signals

    Stars
    3
    Forks
    0
    Open issues
    0
    Weekly downloads
    124
    Last commit
    Mar 3, 2026 (7 months ago)
    Last release
    Mar 3, 2026 (7 months ago)
    Last package publish
    Oct 16, 2026 (12 days ago)

    From GitHub (fetched Oct 24, 2026) and the package registry.

    Details

    Registry name
    io.github.romainsantoli-web/memory-os-ai
    Version
    3.0.1
    Auth
    No auth
    Packages
    pypi memory-os-ai
    License
    GPL-3.0
    First seen
    Oct 24, 2026
    Last checked
    Oct 24, 2026

    More from romainsantoli-web

    Cite this page

    Memory OS AI MCP server. Agent Mart. https://agentmart.dev/connectors/romainsantoli-web-memory-os-ai (last updated 2026-10-24)