EleSync
EleSync is an MCP server that gives AI assistants persistent, shared access to a local markdown vault for reading, writing, and managing memories across clients like Claude, ChatGPT, and Gemini. With it, you can:
Search memories: Find information using hybrid keyword/semantic search, filtering by source, type, and scope.
List and search projects: View available project workspaces and recall memories within a specific project.
Store and update memories: Save facts, preferences, or project state with custom content, tags, types, scopes, and optional templates; mark new memories as superseding old ones to track changes.
Delete memories: Remove a memory by its ID or prefix.
Get vault summaries: See totals of memories by source and type.
Manage scopes: List accessible memory compartments and their counts.
Detect conflicts: Surface contradictory facts from different AI sources.
View history: Examine the full supersession chain of a memory—every prior version and when it changed.
Audit operations: Review all read, write, and delete activity when audit logging is enabled.
EleSync
Your private AI workspace
EleSync gives you a private workspace that your AI tools can share.
When every AI assistant keeps a separate history, you repeat the same background, preferences, decisions, and project details over and over. EleSync keeps the durable context you choose in a local vault, then lets approved assistants use that context with explicit project and scope boundaries.
Start a project with one assistant, continue it with another, and keep the context in a folder you control.
Your local vault
↓
Memories and imported history
↓
Projects, knowledge, tasks, decisions, and handoffs
↓
Approved AI assistantsNo EleSync account, hosted cloud database, e-mail registration, telemetry, subscription, or mandatory cloud sync is required.
Related MCP server: auxly-memory-cli
The five-minute path
A new user should be able to go from installation to a useful AI workspace like this:
Choose a vault. Pick the local folder where EleSync will keep your workspace.
Protect access. Add a local dashboard username and password when you need remote or shared-device access.
Import AI history. Bring in ChatGPT, Claude, Gemini, WhatsApp, or other supported exports and review what will become durable context.
Create a project. Give your memories a purpose and a boundary.
Connect an assistant. Connect Claude Desktop locally, or use a secure HTTPS/OAuth flow for a remote assistant such as ChatGPT, Gemini, or Grok.
Test recall. Ask your assistant:
What useful context can you retrieve from EleSync?
The dashboard includes a guided first-launch wizard and progress tracking for these steps.
What EleSync does
Memory: gives assistants continuity across conversations and providers.
Projects: give context purpose, organization, and boundaries.
Knowledge: stores reusable knowledge sheets, tasks, decisions, and handoffs.
MCP/connectors: let approved AI assistants participate in the workspace.
Imports: bring existing AI history, documents, and WhatsApp conversations into one place.
PWA/mobile: capture and access the workspace from a phone while your computer runs EleSync.
Privacy: keeps the vault local-first and makes sharing explicit.
Important: when you connect an AI assistant, that assistant may read the context you make available to it. Use project boundaries, scopes, read-only permissions, and separate vaults when you need stronger limits.
Visual tour
The dashboard is the center of the workspace: memories, projects, imports, connection health, and mobile access are all reached from one place.
The product workflow is:
Area | What you use it for |
Import | Upload ChatGPT, Claude, Gemini, WhatsApp, JSON, CSV, Markdown, TXT, or ZIP exports. |
Memories | Search, review, edit, and organize imported and saved context. |
Projects | Collect memories, knowledge sheets, tasks, decisions, and handoffs. |
Connect AI | Configure local Claude Desktop or secure remote AI connections. |
Connection Health | See whether the vault, dashboard security, local clients, and remote setup are ready. |
PWA/mobile | Capture text, links, and imports from a phone. |
Privacy by design
Local-first: notes and the search index live in the vault on your computer.
User-controlled: you choose the vault folder, backups, and what becomes durable.
No mandatory cloud account: normal local use works without an EleSync cloud service.
Optional encryption: encrypted vaults use Argon2id and XSalsa20-Poly1305 at rest; plaintext remains the default for portability.
Scoped AI access: client grants can limit which memory scopes an assistant can read or write.
Project isolation: project-bound connections cannot recall or modify memories outside their project.
Reviewable imports: imported conversations and WhatsApp messages can be reviewed before saving.
Install
Windows: easiest path
Download the latest Windows installer, run it, and open EleSync when installation finishes.
The installer asks where to keep the vault and visibly lists the AI assistants available from Connect AI after installation:
ChatGPT · Claude · Gemini · Grok · Perplexity · Copilot · DeepSeekClaude Desktop is the optional local connection that can be pre-wired during installation. Remote assistants require your authorization and a secure HTTPS connection, so the installer does not silently configure them.
Windows: Scoop
scoop install https://raw.githubusercontent.com/darknodebros/EleSync/master/elesync.jsonUpdate later with:
scoop update
scoop update elesyncPython / pipx
For macOS, Linux, or users who prefer Python tooling:
pipx install elesync
ele webRequires Python 3.10 or newer. You can also use pip install elesync.
Standalone builds
Standalone Windows, macOS, and Linux builds are attached to each GitHub release.
First import
From the dashboard:
Open Import.
Choose a supported file or drop it into the import area.
Review the detected format and privacy information.
For WhatsApp, filter by sender and approve only the messages you want to keep.
Save the full conversation or selected durable memories.
Open Memories to search what was imported.
Supported formats include:
JSON · CSV · TXT · Markdown · ZIP · WhatsApp exports · Grok exportsImports are local. Nothing becomes durable project knowledge until the user saves or approves it.
First project
Open Projects, create a project, and give it a short purpose such as:
Japan trip planning
Product launch
Biology assignment
EleSync workspaceInside a project you can:
Add existing memories
Save knowledge sheets
Track tasks
Record decisions
Leave AI handoffs
Set AI read/write/conversation permissions
Generate a project-specific local Claude connection
First AI connection
Claude Desktop: local project connection
From a project page, choose the Claude provider and select Generate project connection. EleSync creates a command like:
ele --vault "D:\notes" onboard --target claude-desktop --project PROJECT_ID --name elesync-project-a1b2c3d4Run the command, restart Claude Desktop, and use the entry named:
elesync-project-a1b2c3d4This entry is project-bound. Memories saved through it are attached to the selected project.
It can coexist with a general elesync entry; remove or disable the general entry when you
need strict project-only visibility.
ChatGPT, Gemini, Grok, and other remote assistants
Remote assistants require a public HTTPS origin and provider authorization. Use the Connect AI page to generate the current setup instructions. EleSync will show:
MCP server URL
Secure authentication details
Local server command
Tunnel/reverse-proxy target
Connection verification result
Do not use 127.0.0.1, a private LAN address, or the dashboard port as a public provider URL.
Mobile companion (PWA)
EleSync can be installed on a phone without a separate mobile app:
Start EleSync on your computer with
ele web.Open Use on another device in the dashboard.
Put your phone and computer on the same Wi-Fi network.
Open the displayed address on your phone.
Choose Add to Home Screen or Install app.
The computer must be running for live vault access. The PWA supports shared text and link capture, encrypted offline drafts, automatic retry, and mobile imports. Do not expose the dashboard publicly without HTTPS and a passcode.
Common commands
ele web # open the local dashboard
ele doctor # check installation health
ele import memories.json # import exported memories or notes
echo "I prefer dark mode" | ele import -
ele remember "Prefer short answers with examples"
ele recall "how should my AI answer me?"
ele sync # rebuild the local search index
ele summary # show a vault summaryAdvanced commands cover encryption, backups, USB vaults, scopes, conflict resolution, HTTP serving, and client wiring. See SETUP.md for the full walkthrough.
Troubleshooting
Run:
ele doctorCommon fixes:
If the dashboard does not open, run
ele weband copy the error.If an AI assistant cannot see memories, restart it after setup.
If search looks stale, run
ele sync.If an import looks wrong, keep the original export and check its source format.
If a remote connector is rejected, verify HTTPS, the public
/mcpURL, the tunnel target, and the authentication state in Connection Health.
Documentation
Project status
EleSync is at 1.24.18 (904 tests, ~12,500 LOC). The current product focus is stabilization: bug fixing, UX polish, documentation, first-run completion, privacy clarity, and reliability.
License
MIT. See LICENSE.
Available Tools
12 toolsaudit_logA
View the audit trail of operations on this vault. Shows who did what
and when — useful for reviewing which AI clients are reading or writing
memories. Audit logging is enabled by default; set ELESYNC_AUDIT=0 to disable it.
last limits the number of entries (newest first).
operation filters to a specific op (REMEMBER, RECALL, FORGET, SCOPE, STATUS).
client filters to a specific client label.
Inputs: last, client, operation. Returns: result
find_conflictsA
Surface facts in the vault that look like they're about the same thing but disagree (e.g. two different stated locations), each with its source and date. Use this to spot stale/contradictory memory and confirm with the user.
Inputs: no named parameters. Returns: result
forgetA
Delete a memory by id (full or 8-char prefix shown by recall).
Inputs: memory_id (required). Returns: result
historyA
Show the supersession history of a memory (id or 8-char prefix): every
version that led to or replaced it, oldest first, with when each changed.
This is the "what was true when" view — use it to see what a fact used to
be after a memory was superseded (via remember(..., supersedes=...) or by
resolving a conflict).
Inputs: memory_id (required). Returns: result
list_projectsA
List the user's EleSync projects and their short ids. Use this before
recall_project when the user refers to a project by name.
Inputs: no named parameters. Returns: result
list_scopesA
List the memory scope compartments this client can see, with how many
memories each holds. general is the default, everyone-can-see bucket.
Inputs: no named parameters. Returns: result
memory_statusA
Summary of what's in the vault: totals by source and type (limited to the memories this client's scope can see).
Inputs: no named parameters. Returns: result
project_contextC
Return the approved project knowledge sheets, tasks, decisions, handoffs, and collected memories as a compact context package.
Inputs: project_id (required). Returns: result
recallA
Search the user's unified memory across all AIs. Use this at the start of
a task to recover context. source filters to one AI (chatgpt/claude/gemini);
type filters by fact/preference/project/relationship/conversation/note.
semantic_weight (0.0–1.0) tunes hybrid search: 0.0 = pure keyword,
1.0 = pure semantic, 0.5 = equal. Use -1.0 (default) for the config
default. Results are limited to the memory scopes this client is
allowed to see.
Inputs: query (required), type, limit, source, +1 more. Returns: result
recall_projectA
Search only the chats and memories collected in one EleSync project.
Use the id from list_projects. This is safer and more relevant than a
whole-vault recall for project-specific work.
Inputs: query (required), project_id (required), limit. Returns: result
rememberA
Save a new memory to the user's vault so every other AI can see it too.
Use for durable facts, preferences, project state — not throwaway chatter.
Pass your identity in source (e.g. 'claude', 'chatgpt', 'gemini') so
provenance and conflict detection work correctly across AIs.
scopes (comma-separated, e.g. "legal" or "writing,coding") puts the memory
in a compartment only clients granted that scope can read; omit it to make
the memory general (visible to all) unless this client has a default scope.
supersedes (a memory id or 8-char prefix) marks that this memory replaces
an older one when a fact changes rather than adds (a move, a rename, an
updated preference): the old memory is kept as history but hidden from
recall, and history can then show what changed and when.
template names a structure from the memory://templates resource and
applies its type/tags/scopes defaults; fields is a JSON object filling
that template's placeholders, e.g. {"who": "Ana", "date": "2026-07-25"},
which builds the content for you so repeated captures stay comparable.
Inputs: content (required), tags, type, fields, +4 more. Returns: result
save_project_documentA
Create a new durable document inside one EleSync project.
This is a mutating tool: it writes the title and body to local vault storage
and updates the project manifest. It does not send content to an AI provider,
and it does not overwrite an existing document; each call creates a new
document. project_id must identify an existing project and kind must be
one of knowledge, task, decision, or handoff. Call it only after the
user has approved saving the content. A project-bound connection can use this
only for its assigned project and only when it has project write permission.
Inputs: body (required), title (required), project_id (required), kind. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 12 tools
Each tool has a distinct purpose: project_context vs recall_project, remember vs save_project_document, and memory-specific tools (forget, history, find_conflicts) are clearly separated. No overlaps that could confuse an agent.
Naming is inconsistent: some tools use verb-first patterns (recall, list, save, remember, forget), while others are noun-first or standalone nouns (project_context, memory_status, history, audit_log). The mix of underscore and single-word forms (e.g., list_scopes vs recall_project) further reduces predictability.
12 tools is a reasonable number for a server that combines memory management and project context. It covers the core operations without being overwhelming, though a few could potentially be merged.
Memory operations are well covered (create, read, update via supersedes, delete, search, status, audit). However, project tools lack update/delete capabilities for projects and documents, leaving a notable gap in lifecycle management.
Maintenance
Related MCP Connectors
One memory, every AI. A shared, user-owned markdown memory your AI clients read and write over MCP.
Shared AI memory. ChatGPT, Claude, Cursor and any MCP app read and write the same memory.
Notes and memory assistant: one Markdown vault that Claude, ChatGPT and any MCP client can share.
Persistent memory for AI assistants. Save once; recall from Claude, ChatGPT, or any MCP client.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceA local-first CLI and MCP server that helps you build and search a personal knowledge vault from Markdown notes, with semantic search and AI-powered features like stale note detection and session memory harvesting. It’s provider-agnostic, requires no GPU in its default mode, and exposes your vault as long-term memory to any MCP-compatible AI tool like Claude Code.47Apache 2.0
- AlicenseNot gradedqualityBmaintenanceLocal-first, file-based memory layer for AI agents — one shared Markdown vault across Claude, Codex, Gemini, Cursor and any MCP client. Provides read/write memory tools with an audit trail, per-agent trust levels, and Git sync; no cloud and no lock-in.2MIT
- FlicenseNot gradedqualityCmaintenancePortable AI memory vault — save memories, skills, and configs to a local database and share them across Claude Desktop, Claude Code, Cursor, Cline, and more.1-
- AlicenseAqualityBmaintenanceLocal-first memory engine for AI-agent teams: private/team/project ACL, associative recall, and federated sync across nodes. One SQLite file, no LLM required.129Apache 2.0