Skip to main content

You

Your personality file, per-domain autonomy, capture sources, and what Open Brain does with your data.

Account

You are signed in to Open Brain on this device. Manage how you sign in, switch identity, or sign out of every surface from the dedicated account page.

Your personality file

Free-text describing who you are and how you want LLMs to respond. Saarthi reads this on every call. For ChatGPT, Perplexity, Gemini and other apps that don't speak MCP, hit Copy and paste into the app's custom-instructions slot.

0 characters
Saved alongside agent-managed sections: [CURRENT FOCUS], [BEHAVIORAL PROFILE], [CAPABILITY PROFILE], [FEEDBACK PROFILE]. Saving here only replaces the [USER PERSONALITY] block; the agent's own learning sections are preserved.

Autonomy by domain

Per-domain trust levels Saarthi currently operates at. Promote a domain to widen what it can do without asking; lock a domain to freeze auto-graduation and auto-demotion.

Loading…
L0 Observer · ask first
L1 Assistive · notify
L2 Supervised · auto + log
L3 Autonomous · proactive
L4 Self-Improving · goal-directed
L5 Self-Directing · initiates workflows

Voice & style

How you write and how you think. Open Brain injects this into everything it writes for you (journal drafts, briefs, chapters, email replies, chat answers, notes) so it all reads in your voice, not a generic AI one. Let it learn from your own writing, then edit anything.

Keep a profile per mode (e.g. professional and creative). The active profile is the one Open Brain writes in. You can also read + refine these from Claude/Cursor over MCP (get_personal_context, update_writing_style).

Capture sources

open full page ↗

Your data

open full page ↗