24 connectors. One private brain.

Data we leverage

Your LinkedIn data export (the full archive): profile, positions, education, skills, connections, recommendations, endorsements, posts, comments, job applications, saved jobs, LinkedIn's inferences about you, and your search history. Messages contribute only a per-person frequency signal — the text is never read.

What we extract

• Identity + full career history (positions, education, skills, certifications)
• Every connection, merged with the same person seen in other sources
• Recommendations & endorsements — your reputation layer
• Your posts and comments — the voice layer
• Job applications, saved jobs and reusable application answers
• LinkedIn's algorithmic inferences and ad profile — the mirror layer
• Search history

Where it maps in the brain

Get the export

1. linkedin.com → Me → Settings & Privacy → Data privacy
2. Get a copy of your data → choose the larger data archive (everything)
3. A partial archive arrives in ~10 minutes; the full one within 24 h — use the second email's link (valid ~72 h)

Import folder: linkedin — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Message text is never read — only who + when, as a relationship-strength signal.

Data we leverage

Your Facebook information download (JSON): friends, followers and requests, posts, comments and reactions, liked pages and saved items, check-ins and cities, ad-interests and off-Meta activity, events with your RSVP, notes, page reviews and marketplace listings.

What we extract

• Friends, followers and requests — your people graph
• Posts, comments and reactions — the voice layer
• Liked pages and saved items — interests
• Check-ins and cities — geocoded map places
• Ad-interests, advertisers and off-Meta activity — Meta's algorithmic mirror of you
• Events with your RSVP

Where it maps in the brain

Get the export

1. accountscenter.facebook.com → Your information and permissions
2. Download your information → Facebook profile
3. Format JSON · media quality low · date range All time — ready within hours

Import folder: facebook — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Message bodies are never read — signal only.

Data we leverage

Your Instagram download (JSON): followers/following, posts and comments, topics and liked/saved content, last-known location and photo EXIF places, and story interactions (polls, quizzes, likes) as activity counts.

What we extract

• Followers and following — your people graph
• Posts and captions, your comments — the voice layer
• Topics and liked/saved content — interests
• Photo EXIF and last-known location — map places
• Story interaction counts

Where it maps in the brain

Get the export

1. accountscenter.facebook.com → Download your information
2. Choose the Instagram profile → Format JSON

Import folder: instagram — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

DM text contributes only a per-person frequency signal.

Data we leverage

Your Google Takeout: Contacts, Calendar, Maps saved/reviewed/labeled places, Semantic Location History visits, photo-spot pins from EXIF sidecars, Search & Ads My Activity, YouTube subscriptions/watch/search/likes/comments, and Chrome bookmarks. Gmail, raw GPS pings, Fit, Drive and Keep are excluded by design.

What we extract

• Contacts — the canonical people list
• Calendar — events, with who you met
• Maps places, visits and photo spots — a geocoded map of your life
• Search & Ads activity — search history + the algorithmic mirror
• YouTube taste graph — subscriptions, watch and search history
• Chrome bookmarks — interests

Where it maps in the brain

Get the export

1. takeout.google.com → Deselect all
2. Select Contacts, Calendar, Maps (your places), Saved, Location History / Timeline, My Activity, Chrome, YouTube, Google Photos
3. Export once — large exports arrive as several "Takeout N" zips over hours–days; drop them all in together

Import folder: google — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Gmail, Fit, Drive and Keep are excluded by design and reported as skipped.

Data we leverage

Your X archive: tweets and long-form note-tweets (retweets excluded), following and followers, lists, and like counts. DMs and the imported address book are quarantined — never read.

What we extract

• Tweets + note-tweets — the richest public-voice archive most people own
• Following and followers — your people graph (handles)
• Lists — interests
• Like counts

Where it maps in the brain

Get the export

1. x.com → Settings → Your account → Download an archive of your data
2. Re-verify your identity when asked
3. Wait 24–48 h for the email, then download the ZIP

Import folder: x — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

DMs and the imported address book are quarantined — never read.

Data we leverage

Exported chats (without media): who you talk to, how often, how recently. The parser stops at the sender — message text is never read.

What we extract

• Contact graph — the people you actually talk to
• Per-person message frequency and recency — relationship strength

Where it maps in the brain

Get the export

1. Open a chat → ⋮ → More → Export chat → Without media
2. Save the .txt — repeat for the chats that matter

Import folder: whatsapp — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Signal-only by design: message text is never read.

Data we leverage

Your GitHub account export: repositories with descriptions and issue/PR activity counts, languages and starred repos, followers/following, and organization memberships.

What we extract

• Repositories — code voice, with activity counts
• Languages and stars — technical interests
• Followers and following — the dev network
• Organization memberships

Where it maps in the brain

Get the export

1. github.com → Settings → Account → Export account data → Start export
2. Download from the email link (arrives within hours)
3. Extract the tar.gz before importing

Import folder: github — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

The YouTube slice of Google Takeout: watch history, search history, subscriptions, playlists (with per-playlist video titles), and your comments.

What we extract

• Watch history — a channel-level taste graph
• Search history
• Subscriptions and playlists — interests
• Your comments — voice

Where it maps in the brain

Get the export

1. takeout.google.com → Deselect all → tick YouTube and YouTube Music only
2. In options keep history / subscriptions / playlists / comments — untick videos

Import folder: youtube — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

Your Strava bulk export: activities, clubs, follower/following athletes, and your profile city. Each activity contributes its start point only — full routes are never imported.

What we extract

• Activities — events + sport interests
• Training spots — map pins from activity start points only
• Clubs and fellow athletes

Where it maps in the brain

Get the export

1. strava.com → Settings → My Account → Download or Delete Your Account
2. Download Request → ZIP arrives by email

Import folder: strava — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Routes are reduced to start points — never the full track.

Data we leverage

Your Reddit GDPR export: posts and comments, subscribed subreddits and multireddits, friends, saved posts, and search history. Private messages contribute sender + date signal only.

What we extract

• Posts and comments — candid voice
• Subreddits and multireddits — interests
• Friends, saved posts, search history

Where it maps in the brain

Get the export

1. reddit.com/settings/data-request → request the full GDPR export
2. Download from the email link (typically under 48 h)

Import folder: reddit — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

PM subjects and bodies are never selected — signal only.

Data we leverage

Your Spotify download: listening history, followed artists, playlists and their tracks' artists, Spotify's Inferences about you, Marquee ad segments, and search queries. Userdata and payment files are quarantined.

What we extract

• Listening history — an artist-level taste graph
• Playlists and followed artists
• Spotify's Inferences and Marquee segments — the algorithmic mirror
• Search queries

Where it maps in the brain

Get the export

1. spotify.com/account → Security & privacy → Download your data
2. Tick Extended streaming history — the default alone is thin
3. Email links arrive separately (account ~5 days, extended up to 30) — drop both zips' contents together

Import folder: spotify — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Userdata and payment files are quarantined, never imported.

Data we leverage

Your TikTok data download (JSON): following, searches, favorite hashtags, your comments, and like/favorite counts. DM text sits in the same file but no rule ever selects it.

What we extract

• Following — people
• Searches and favorite hashtags — interests
• Your comments — voice
• Like/favorite counts

Where it maps in the brain

Get the export

1. App → Profile → ☰ → Settings → Account → Download your data
2. Choose Format: JSON (machine-readable)
3. Ready in days — download within 4 days

Import folder: tiktok — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

DM text is in the file but no mapping rule ever selects it.

Data we leverage

Page-admin exports: the org profile (HQ geocoded), the employee list with departments, followers, page posts, and follower/visitor analytics — the company's algorithmic mirror.

What we extract

• Org profile with geocoded HQ
• Employees with departments — org structure
• Followers and page posts — company voice
• Page analytics — the algorithmic mirror

Where it maps in the brain

Get the export

1. Page admin → Analytics / Settings
2. Export the employee list, followers, posts and analytics reports (per-report CSVs — no single ZIP)

Import folder: linkedin_company — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

Admin exports: users with Org Units, shared calendars with attendees, shared drives, and groups. Login/admin audit logs are quarantined.

What we extract

• Users with Org Units — employees + department structure
• Shared calendars with attendees — the meeting graph
• Shared drives — projects
• Groups — teams

Where it maps in the brain

Get the export

1. admin.google.com → Directory → Users → Download users (CSV)
2. Calendars / drives via org-wide Takeout or Google Vault

Import folder: google_workspace — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Login and admin audit logs are quarantined.

Data we leverage

The Slack workspace export: members, channels with topics and membership, and per-person message frequency. Message text is never read.

What we extract

• Members — people
• Channels with topics + membership — the who-knows-what map
• Per-person message frequency — collaboration signal

Where it maps in the brain

Get the export

1. Workspace admin → <workspace>.slack.com/services/export
2. Export (public channels; available on all paid tiers)
3. Private channels / DMs need an Enterprise Grid compliance export

Import folder: slack — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Message text is never read — only who + when.

Data we leverage

The workspace export (Markdown & CSV): every wiki page with breadcrumb context and intra-wiki references, plus database rows.

What we extract

• Wiki pages — institutional voice, with "Projects / Roadmap — …" breadcrumbs
• Database rows — interests / structured records

Where it maps in the brain

Get the export

1. Workspace owner → Settings → Export all workspace content
2. Markdown & CSV, include subpages → download from the email link

Import folder: notion — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

A space export (XML for full fidelity; HTML also parses): pages with their authors, the space itself, and body excerpts.

What we extract

• Pages with authors — expertise signal (who wrote what)
• Spaces — org structure
• Body excerpts — institutional voice

Where it maps in the brain

Get the export

1. Space admin → Space settings → Export space
2. Choose XML (full fidelity) → ZIP

Import folder: confluence — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

Issue CSV exports: projects, assignees and reporters with per-project ownership tags, components and labels, and comment authors. Ticket prose is never imported.

What we extract

• Projects — orgs
• Assignees / reporters — people with project ownership tags
• Components & labels — the technology map
• Comment authors — who-answers-what signal

Where it maps in the brain

Get the export

1. Issue navigator → filter the project(s)
2. Export → CSV (all fields) — repeat per 1000-row page if large

Import folder: jira — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Ticket prose is never imported — structure and ownership only.

Data we leverage

The Data Export Service ZIP: accounts (geocoded), contacts and leads joined to accounts, opportunities, campaigns with members, tasks and cases as per-contact interaction signal, and users (your own team). Subjects are never read.

What we extract

• Accounts — customer orgs, geocoded
• Contacts & leads — people joined to their accounts
• Opportunities — the deal timeline (one note per deal)
• Campaigns + members
• Tasks & cases — per-contact interaction signal

Where it maps in the brain

Get the export

1. Setup → Data Export Service → Export Now (or schedule)
2. Download from the email — the link expires in 48 h → CSV ZIP(s)

Import folder: salesforce — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Task/case subjects are never read.

Data we leverage

Per-object CSV exports: companies (geocoded), contacts, deals, tickets as per-contact support signal, and deal/ticket owners (your own team).

What we extract

• Companies — customer orgs, geocoded
• Contacts — people
• Deals — the pipeline timeline
• Tickets — per-contact support signal
• Owners — your own team

Where it maps in the brain

Get the export

1. Per object: Contacts / Companies / Deals / Tickets list → Export
2. CSV, all properties — keep the hubspot-… filenames

Import folder: hubspot — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Data we leverage

The account data export: customer organizations, requesters joined to their org, and ticket volume with per-requester signal. Subjects and descriptions are never read.

What we extract

• Customer orgs
• Requesters — people at their org
• Ticket volume + per-requester signal — product pain map

Where it maps in the brain

Get the export

1. Admin Center → Account → Requests to export data
2. (Zendesk support must enable the export once) → JSON/CSV by email

Import folder: zendesk — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Ticket subjects and descriptions are never read.

Data we leverage

An .mbox archive (Gmail Takeout / Google Vault; Outlook PST converted first): every correspondent's display name, with frequency and recency. Headers only — subjects and bodies are never accessed.

What we extract

• Correspondents — people
• Per-person frequency + recency — relationship strength

Where it maps in the brain

Get the export

1. Gmail: Takeout → Mail → MBOX (or Google Vault org-wide)
2. Outlook PST: convert first — readpst -o out/ archive.pst — then drop the .mbox

Import folder: email — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Headers only — subjects and bodies are never accessed.

Data we leverage

Purview eDiscovery results (.eml + results CSV): the same headers-only people + relationship signal as the mbox source.

What we extract

• Correspondents — people
• Per-person frequency + recency — relationship strength

Where it maps in the brain

Get the export

1. Purview compliance portal → eDiscovery → Content search
2. Export results (requires the eDiscovery Manager role)

Import folder: microsoft365 — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Headers only — bodies never accessed.

Data we leverage

A Purview eDiscovery message report: senders, teams and channels with membership tags, and message frequency. Content columns are never read.

What we extract

• Senders — people
• Teams & channels — orgs with membership tags
• Message frequency signal

Where it maps in the brain

Get the export

1. Purview eDiscovery → message report (CSV/JSON) for the teams in scope

Import folder: teams — or just tell Claude Code / Codex “build my brain from this export”, or drop the .zip on the connector tile in Studio.

Content columns are never read — signal only.