SULCUS

Memory That Thinks.

The Proactive Memory Layer for AI Agents

AI agents forget. Context windows fill up, and simple search returns irrelevant noise. Sulcus gives your agents persistent, self-updating memory. It surfaces the right background before every response and learns continuously as they work.

Autonomous Memory

Proactive Context. Zero Boilerplate.

Most memory tools are passive. They sit idle until an agent remembers to call them. Sulcus runs proactively. It surfaces the right context automatically before your agent answers, and saves key insights when it finishes.

  • Automatic Recall — Relevant memories, user preferences, and project facts appear automatically before your agent answers.
  • Continuous Learning — Captures new insights and user preferences in the background without extra prompts or token waste.
  • Connected Knowledge — Links people, tools, and decisions together so your agent understands the complete picture.
  • Self-Cleaning Context — Important facts stay sharp forever, while temporary conversation noise fades out automatically.
  • Smart Summarization — Older interactions condense into compact summaries, cutting prompt token costs by up to 90%.

Built For Production Agents

Four foundational layers that make agent memory dependable.

01

Intelligent Extraction

Every interaction is analyzed automatically. Facts, preferences, and key instructions are extracted and indexed while conversational noise is discarded.

02

Connected Knowledge

Memories connect across people, projects, and tools. When your agent recalls a topic, it gets the full context instead of isolated fragments.

03

Adaptive Retention

Critical preferences and instructions stay permanently accessible. Fleeting details naturally fade out so prompt windows never get bloated.

04

Memory Compression

Past sessions condense into compact summaries in the background. Cut token costs by up to 90% while keeping full conversational continuity.

Complete Control

Fully Configurable. No Black Boxes.

Tune context budgets, recall thresholds, automated actions, and retention schedules via simple API settings.

Active Index
max_nodes
context_budget
hot_threshold
cold_threshold
Resonance
spread_factor
depth
damping
filter_gate
Reinforcement
on_recall
on_update
on_edge
stability_gain
Consolidation
cold_trigger
cold_threshold
strategy
Engine Ticks
interval_ms
trigger_ops
max_idle_ms
Retention Profiles
half_life
floor
reinforce
stab_gain

Automation First

Reactive Memory Triggers

Set automated rules on your memory graph. Pin critical facts, reinforce frequently accessed knowledge, and send webhooks to your external tools without manual prompting.

on_store→pin

Pin preference memories automatically so they never fade away.

on_recall→boost

Reinforce memories whenever they are searched so important knowledge stays sharp.

on_decay→notify

Send an alert when key knowledge begins cooling down before it is lost.

on_boost→webhook

Trigger external webhooks when important memories are reinforced.

4 events · 7 actions · filter by type, namespace, or custom rules · unlimited triggers

Developer-First

Five lines. Real memory.

Zero-dependency SDKs for Python and Node.js. Or connect directly via MCP or REST.

Pythonpip install sulcus
from sulcus import Sulcus

s = Sulcus(api_key="sk-...")
s.remember("User prefers dark mode",
           memory_type="preference")
results = s.search("dark mode")
Node.jsnpm install @digitalforgestudios/sulcus
import { Sulcus } from "@digitalforgestudios/sulcus";

const s = new Sulcus({ apiKey: "sk-..." });
await s.remember("User prefers dark mode",
  { memoryType: "preference" });
const results = await s.search("dark mode");
Full Documentation →

Privacy-First Architecture

Your memories. Your machine. Your rules.

Most memory systems require your data to live on someone else's server. Sulcus runs locally by default. Cloud sync is optional, encrypted, and you control what leaves the machine.

01

Local-First

The core engine runs entirely on your local hardware. No network calls are required to read or write memories.

02

Zero Knowledge Sync

Cloud sync is encrypted in transit and at rest. Strict tenant isolation ensures your memory graph remains completely private.

03

Data Sovereignty

Self-host the entire stack in your own private cloud or data center with no tracking and full enterprise control.

04

Selective Sync

Choose exactly which workspaces sync to the cloud and which stay strictly on your local device.

GDPR-ready · SOC2 roadmap · No telemetry · Enterprise on-premise available

Performance Validated

<25ms
Context Build Time
87
API Endpoints
v2.9
Current Release
15k
Knowledge Graph Vertices
View Documentation →

Give Your Agents a Memory.

Free tier. No credit card required. Start building agents with persistent memory in under five minutes.

Create Free Account

No credit card required · Privacy-first · Full data control

Stay in the Loop

Releases, memory research, and what we're building.