DanClaw Massive Memory Brain

6-Layer Persistent Intelligence System

~4k tokens startup load
L1
0
Session Logs
entries today
L2
~3k
Memory File
tokens
L3
1
Projects
registered
L4
0
Vectors
stored
L5
~80%
Token Save
efficiency
L6
Intelligence
last run

6-Layer Memory Architecture

Live
Layer 1

Session Logs

YYYY-MM-DD.md
On-demand load

Layer 2

Long-Term Brain

MEMORY.md
Loads at startup

Layer 3

Project Registry

projects.md
Loads at startup

Layer 4

Vector Memory

Semantic search
On-demand recall

Layer 5

Smart Loading

~4k token startup
80% savings

Layer 6

Neural Engine

INTELLIGENCE.md
Pattern analysis

Startup: L3 projects.md (1k) + L2 MEMORY.md (3k) = ~4k tokens  ·  Session logs + Vector search = on-demand only  ·  Neural engine = 24h cycle

Latest Intelligence

Run neural analysis to generate insights...

Memory Token Budget

projects.md (L3) 15%
MEMORY.md (L2) 45%
INTELLIGENCE.md (L6) 20%
Session logs (on-demand) 10%
Vector results (on-demand) 10%
1
Session Intelligence

Daily Session Logs

Automatic daily archives of all agent interactions · memory/YYYY-MM-DD.md

No session entries yet.

Use agents in the dashboard — entries are automatically written here.

Write Manual Entry

2
Long-Term Brain

MEMORY.md

Core intelligence file · Loaded at every agent startup

tokens
sections
characters

Append to Section

3
Project Registry

projects.md

Instant situational awareness · Loaded at startup (~1k tokens)

4
Semantic Vector Memory

Search by Meaning

Retrieve relevant memories using natural language — not keywords

Enter a query to search semantic memory.

Memory grows with every agent interaction.

Store Memory Manually

5
Smart Context Loading

Token-Efficient Startup

Load only what agents need. Reduce context from 20k+ → ~4k tokens.

Loaded at Startup

projects.md

Project registry with status and context

~1k

MEMORY.md

Strategic decisions, rules, core knowledge

~3k

Loaded On-Demand

Daily session logs

Only when historical context needed

variable

Vector search results

Only when semantic recall needed

top-5 only

INTELLIGENCE.md summary

Injected as compressed briefing

~0.5k

Live Context Snapshot

Click refresh to load context snapshot...

Build Enriched Agent Prompt

6
Neural Intelligence Engine

INTELLIGENCE.md

Pattern analysis · Opportunity detection · Strategic insights · 24h cycle

Trends
Opportunities
Risks
Actions

Neural engine ready.

Click "Run Analysis Now" to generate intelligence from your memory base.

Auto-Curation Engine

Runs automatically every 3.5 days · Keeps MEMORY.md accurate and lean

1

Read Session Logs

Scans recent daily logs for high-value entries

2

Update MEMORY.md

Extracts decisions and curated insights

3

Update Projects

Refreshes project statuses and activity

4

Remove Outdated Info

Prunes stale data to keep context tight

5

Flush to Vector Store

Embeds important entries for semantic recall

6

Trigger Neural Run

Updates INTELLIGENCE.md with fresh analysis

Run Curation Now

Auto-runs every 3.5 days. Manual trigger available.

Write to Memory

Manually inject knowledge into any memory layer

Append to MEMORY.md

Store in Vector Memory