Your social history, made usable by agents. -> mymem.space
MyMem at mymem.space turns the social media exports you already own into one searchable memory. The deployed web app gives you a visual timeline across platforms, lets you search your own history, and gives an agent the context of what you actually posted.
You do not need to install software or write code to use the deployed product. Start at mymem.space.
Ask MyMem questions such as:
"What have I said about facial recognition?"
"Which restaurants have I eaten at in Mumbai?"
"Where did I buy a Thanka painting in Kathmandu?"
The archive answers from your imported history with dates and source links, newest first.
The archive browser: one timeline across 149,950 items, with platform and activity filters.
Why use MyMem at mymem.space?
Platform search is built for finding public content, not for finding the thing you said three years ago. MyMem searches across X, LinkedIn, Instagram, and Facebook exports in one place. It combines SQLite FTS5 keyword search with local embeddings, so it finds both the words you used and the meaning behind a query.
Your archive is a copy of your own social history. If a platform locks your account or changes its search, your exported posts, replies, likes, saves, follows, and connections remain available in MyMem.
An agent queries the deployed archive over MCP and answers from what you actually posted. It finds your past positions, saved items, people, and timelines instead of starting from a blank context.
The payoff: Saki answers from the owner's actual 2015 posts.
The web archive gives you a visual timeline of your social media across platforms. Posts, replies, likes, saves, and other public activity stay connected to their platform, date, and link.
Use the deployed product at mymem.space
Use each platform's normal data-export page. Select JSON and the full date range when those choices are offered.
- X: Settings and privacy -> Your account -> Download an archive of your data.
- LinkedIn: Settings and Privacy -> Data privacy -> Get a copy of your data.
- Instagram: Accounts Center -> Your information and permissions -> Download your information. Choose JSON and all time.
- Facebook: Accounts Center -> Your information and permissions -> Download your information. Choose JSON.
- YouTube and Google activity: Use Google Takeout and select YouTube and My Activity. Choose JSON.
Each platform gives you a ZIP file. Keep the original files. MyMem reads the exports and does not need live platform APIs or OAuth connections.
2. Upload at mymem.space
Open mymem.space, choose the archive upload control, and select one export ZIP. Upload each platform's ZIP separately. Large exports upload in chunks. The page reports the import job and item counts when parsing finishes.
The public import filters the export to public content. Direct messages and private profile information do not reach the published archive. MyMem processes the ZIP and deletes it after it is read.
Open the archive browser at mymem.space after the import completes. Use the per-year timeline to move through your history. Filter by platform, then narrow by kind such as post, reply, like, connection, watch, save, or comment. Open an item to follow its source link when the export includes one.
The published archive keeps public likes, saves, follows, and connections because they are part of the public social record represented by the exports. Reposts and retweets are credited to the original author, not to you. Quote-tweets keep your commentary as your own text and retain the quoted post as a link.
4. Connect an agent to mymem.space
Agents use the read-only MCP endpoint at:
https://mymem.space/mcp
The endpoint requires a bearer token. Put the token supplied to you in the Authorization header:
mcp_servers:
mymem:
url: "https://mymem.space/mcp"
headers:
Authorization: "Bearer YOUR_MyMem_TOKEN"Replace YOUR_MyMem_TOKEN, then reload your MCP client. Claude Code uses the same deployed endpoint:
claude mcp add --transport http mymem https://mymem.space/mcp --header "Authorization: Bearer YOUR_MyMem_TOKEN"The endpoint serves the published copy and does not expose direct messages. See docs/REMOTE.md for the endpoint contract.
Only public content is imported into the public archive. Direct messages and private profile information never reach it. Public likes, saves, follows, and connections remain because they are part of the public social record represented by the exports.
Reposts and retweets are never attributed to you as something you said. They are credited to the original author. Quote-tweets keep your words as your own commentary; the quoted post is reduced to a link.
The working archive and published archive stay separate. The web viewer and MCP endpoint serve only the filtered published copy.
- Request data exports from the platforms you use.
- Upload the ZIP files at mymem.space.
- MyMem parses each platform's export, filters the public archive, and builds the search index.
- Browse the timeline at mymem.space or connect an agent to
https://mymem.space/mcp.
The adapters read each platform's own export. They do not depend on live platform APIs or OAuth connections.
MyMem was built with Agent Orchestrator (AO): one orchestrator session coordinated many worker sessions, with each worker owning a task and its pull request. Roughly 38 worker sessions were spawned over the build, and 22 pull requests were opened.
Workers ran in parallel on independent slices: the remote MCP endpoint, chunked archive uploads, repost attribution, the public-only import filter, the documentation rewrite, the wordmark, homepage styling, and search ordering. The orchestrator did not write feature code. It inspected state, spawned and redirected workers, routed corrections, and merged completed work.
The orchestrator verified the live site instead of relying on worker reports. A production outage also showed why that matters: a schema change reached code before the separately generated published database. Workers repaired the outage and added compatibility for older published schemas.
 Agent Orchestrator merges independent worker sessions and their pull requests. |
The normal user path is the deployed web app at mymem.space. The repository also contains the local engine for development and direct work with an archive.
git clone https://github.com/nixxintools/mymem
cd mymem
npm install
npm test
npm run buildThe local engine can import an export, search it, build a local timeline, and run the MCP server over stdio:
npm run mymem -- import path/to/export.zip
npm run mymem -- search "facial recognition"
npm run mymem -- stats
npm run mcpUse docs/VERIFY.md for contributor verification and docs/REMOTE.md for the deployed MCP contract.
MyMem is open source under the MIT license. The engine, adapters, and web application are in this repository: github.com/nixxintools/mymem.