Welcome
Hi.
Loading your dashboard…
Hello.
Pinned
Loading…
Recent
- Loading…
Loading…
Landing page
Dashboard
Viewing dashboard of —
Memory
What Familiar knows
—facts
—entities
—relationships
No memories yet. Say hi to Familiar to get started.
View memory →
Graph
Entities by type
No entities yet. Conversations about people, places, and projects will fill this in.
View graph →
Live
Recent sessions
No recent sessions. Start a chat to see activity here.
Recent
Saved memories
Familiar hasn't saved anything yet.
View memory →
Agents
Shards
—
You don't have any shards. Create one to expose Familiar as a focused agent.
View all →
Growth
Memory over 30 days
Too early to chart — come back after a few days of activity.
Preview
Top entities & connections
Loading
Your graph is empty. Chat with Familiar to build connections between people, places, and projects.
Pick a section from the left.
System status
Live gateway health, model status, and aggregate store counts. Refreshes every 30 seconds while this panel is visible.
Gateway
Uptime
—
Memory
Total rows
—
Users
—
—
Sessions
Active
—
Models
| ID | Provider | Status | Endpoint |
|---|---|---|---|
| Loading | |||
Maintenance mode
Take the primary model out of rotation and route chat to a slower fallback. A warning banner shows for every user while active; it also engages automatically if the primary goes offline.
Skills
—
Users
| Name | ID | Status | Identities | Created | |
|---|---|---|---|---|---|
| Loading | |||||
Token minted
Copy this once
Familiar stores a bcrypt hash — this plaintext value will not be visible again. Copy it to your caller's secret store now.
Book
Members
Loading…
My skills
Your private markdown skills. Only you can see and use them: flip "Use in chat" to make one available in your normal conversations, or bind it to your shards from a shard's detail page. Write one with New skill — plain markdown instructions the assistant follows when the skill applies. Bundled scripts are never executed.
Loading
New skill
lowercase letters, digits, single hyphens — how the assistant activates it; fixed after creation
in every chat's prompt — say WHAT it does and WHEN to use it
Imported skills
Standard Agent Skills (SKILL.md packages) imported into this instance's library. Imported skills only run inside shards — bind them from a shard's detail page. All packages are unsigned for now; bundled scripts are never executed. Admit new packages with Import skill, or drop a directory into the library folder and Rescan.
Loading
Built-in skills
The capabilities compiled into this gateway — memory retrieval, web search, weather, news, and so on. Each exposes one or more tools the assistant can call. Read-only: built-in skills are configured at the gateway and require a restart to change.
Loading
Import skill
Upload a skill zip or paste a direct zip URL. It's admitted into the instance library for shard use. Preview shows exactly what the package declares before it's admitted; bundled scripts are never executed.
Semantic memory
Viewing memory of —
| Content | Scope | User | Source | Conf | Created | |
|---|---|---|---|---|---|---|
| Loading | ||||||
—
| Entity | Connections | Facts | Last seen |
|---|---|---|---|
| Loading | |||
Store health
Memory graph
Entities (people, places, projects, things mentioned in conversation) and the relationships between them. Click a node or edge for detail — including the supporting fact behind an edge. Double-click a node to grow its neighborhood into the view. Edges can be edited (predicate, confidence) or deleted from the detail panel; that touches graph entries only, the underlying memories stay put.
Viewing graph of —
Loading
Scheduled Actions
Have your Familiar perform actions for you in the background on a scheduled basis. You define the work to do, the data that's in scope, and how Familiar responds.
| Name | Trigger | Envelope | Reports to | State | Last run | Last fired |
|---|---|---|---|---|---|---|
| Loading | ||||||
New Action
Ledger
Run history
| Started | Status | Trigger | Model | Duration | Output / error |
|---|
Profile
Editing personality of —
Display name
Assistant personality
Describe how you'd like your assistant to behave — its tone, voice, and style. This travels with every turn and shapes the personality without touching the system prompt your administrator controls. Facts about you (your role, preferences, projects) don't go here — the assistant learns those into its memory as you talk.
System prompt
This is the base instruction set your administrator configured for the assistant. It's read-only — shown here for transparency.
Passkeys & security keys
Add another credential so you can sign in from a different device. Your browser will offer a passkey (Touch ID, Windows Hello, iCloud Keychain) or a hardware security key — pick whichever matches the device you're on. The new credential is bound to the email on your session.
Your credentials
Domains
System prompt
This is the base layer of the assistant's prompt — shared by every user. The classifier-driven tier overlays and tool-policy layer are still appended automatically on top of whatever you save here. Leave the box empty to fall back to the file-loaded default (base.md).
Saved.