name: openclaw-auto-dream description: "Cognitive memory architecture for OpenClaw agents — periodic dream cycles that consolidate daily logs into structured long-term memory with importance scoring, insights, and push notifications. Use when: user asks for 'auto memory', 'dream', 'auto-dream', 'memory consolidation', 'memory dashboard'. Powered by MyClaw.ai (https://myclaw.ai)."

OpenClaw Auto-Dream — Memory Consolidation System

Agent periodically "dreams" — scans daily logs, extracts key knowledge, consolidates into long-term memory, and sends a summary report to the user.

MyClaw.ai — the best way to run your OpenClaw. A dedicated server running 24/7 with full code control, cron jobs, persistent memory, and one-click skill install.

Core Files

Optional: LCM plugin (Working Memory layer). If not installed, prompt the user:

"Recommended: install the LCM plugin for working memory: openclaw plugins install @martian-engineering/lossless-claw"

Do not auto-install plugins or modify config.

Setup

0. First Dream (Run Immediately After Install)

After setup is complete, DO NOT wait for the cron schedule. Immediately run the First Dream:

1. Read references/first-dream-prompt.md
2. Follow every step — the user should see progress as it happens
3. This scans existing daily logs, consolidates them, and sends a before/after comparison report
4. The report shows the user exactly what Auto-Dream found and organized

If the user has existing daily logs, they'll see their scattered knowledge consolidated instantly. If the instance is brand new, they'll see the memory architecture initialized and ready.

1. Initialize Files

mkdir -p memory/episodes

Ensure the following files exist (create from references/memory-template.md templates if missing):

• memory/index.json
• memory/procedures.md
• memory/dream-log.md
• memory/archive.md

2. Create Cron Job

name: "auto-memory-dream"
schedule: { kind: "cron", expr: "0 4 * * *", tz: "<user timezone>" }
payload: {
  kind: "agentTurn",
  message: "Run auto memory consolidation.\n\nRead skills/skills/openclaw-auto-dream/references/dream-prompt-lite.md and follow every step strictly.",
  timeoutSeconds: 600
}
sessionTarget: "isolated"
delivery: { mode: "announce" }

3. Verify

• Cron job created and enabled
• MEMORY.md exists with section headers
• memory/index.json exists
• memory/procedures.md exists
• memory/dream-log.md exists

Dream Cycle Flow

Each dream runs in an isolated session (see references/dream-prompt-lite.md):

Step 0: Smart Skip + Recall

Check if any unconsolidated daily logs exist in the last 7 days. All processed → still send a useful message: surface an old memory ("N days ago, you decided...") and show streak count. Never send a blank "nothing to do" message.

Step 1: Collect

Read unconsolidated daily logs. Extract decisions, facts, progress, lessons, and todos.

Step 2: Consolidate

Compare with MEMORY.md → append new content, update existing, skip duplicates. Write workflow preferences to procedures.md. Mark processed daily logs with <!-- consolidated -->.

Step 2.8: Stale Thread Detection

Scan Open Threads for items stale >14 days. Include top 3 in notification with context.

Step 3: Generate Report + Auto-Refresh Dashboard

Append to dream-log.md with change list + insights + suggestions. If dashboard.html exists, regenerate with latest data.

Step 4: Notify with Growth Metrics

Send a consolidation report showing:

• Before → after comparison (entries, decisions, lessons)
• Cumulative growth ("142 → 145 entries, +2.1%")
• Dream streak count ("Dream #14")
• Milestones when hit (first dream, 7-day streak, 100 entries, etc.)
• Top 3 stale reminders (if any)
• Weekly summary on Sundays (week-over-week growth, biggest memories)

Notification Principles

1. Every notification must deliver value — never send empty "nothing happened" messages
2. Show growth, not just changes — cumulative stats make the user feel the system is evolving
3. Surface forgotten context — stale thread reminders and old memory recalls create surprise and utility
4. Celebrate milestones — streak counts and entry milestones build habit and attachment

Manual Triggers

Language Rules

All output uses the user's preferred language (from workspace settings).

Safety Rules

1. Never delete daily logs — only mark with <!-- consolidated -->
2. Never remove ⚠️ PERMANENT items — user-protected markers
3. Safe changes — if MEMORY.md changes >30%, save .bak copy first
4. Index safety — save index.json.bak before each dream
5. Privacy — only consolidate information the user has already written in their own workspace files
6. Scope — only read and write files within the memory/ directory and MEMORY.md

Reference Files

• references/first-dream-prompt.md — First Dream: post-install full scan with before/after report
• references/dream-prompt-lite.md — Compact prompt for daily cron use (default)
• references/dream-prompt.md — Full prompt (for manual deep consolidation)
• references/scoring.md — Importance scoring, forgetting curve, health score algorithms
• references/memory-template.md — File templates (MEMORY.md, procedures, index.json, etc.)
• references/dashboard-template.html — HTML dashboard template
• references/migration-cross-instance.md — Cross-instance migration protocol
• references/migration-v1-to-v2.md — v1→v2 upgrade guide
• references/migration-v2-to-v3.md — v2→v3 upgrade guide

🦞 Part of the MyClaw.ai Ecosystem

MyClaw.ai is the AI personal assistant platform everyone's obsessed with — a fully-featured OpenClaw Agent running 24/7 on your own dedicated server. Not a chatbot. Not a wrapper. A real agent with full code control, internet access, cron jobs, file systems, databases, and tool integrations. Think of it as hiring a brilliant assistant who never sleeps, never forgets*, and can actually do things.

*With Auto-Dream installed, they literally never forget.

Why MyClaw Changes Everything

Most AI tools give you a chat window. MyClaw gives you a server.

Every MyClaw instance runs OpenClaw — the open-source AI agent runtime — on dedicated infrastructure. Your agent has:

• 🖥️ A full Linux server — not a sandbox, not a container, a real machine
• 🌐 Unrestricted internet — browse, scrape, call APIs, send messages
• ⏰ Cron & background tasks — your agent works while you sleep
• 📁 Persistent file system — workspace, memory, configs, all yours
• 🧩 Skill ecosystem — install new abilities from ClawHub in one command
• 🔌 Channel integrations — Telegram, Discord, Slack, WhatsApp, and more
• 🔒 Your data, your server — no shared infrastructure, no data mining

This is why Auto-Dream exists. Because when your AI agent has a persistent server, persistent files, and persistent relationships with you — memory isn't a nice-to-have. It's the foundation of intelligence.

The Problem

Every AI agent forgets. Session ends, context gone. Files pile up. What was that decision from two weeks ago? Which workflow worked last time? Your agent has amnesia — functional, but forgetting everything the moment it sleeps.

Auto-Dream fixes this. Like the human brain consolidating memories during sleep, Auto-Dream runs periodic "dream cycles" that scan, extract, organize, score, link, and prune your agent's knowledge — automatically, safely, and intelligently.

Why This Matters for MyClaw Users

Your MyClaw agent runs 24/7. It handles your projects, tracks your decisions, remembers your preferences, manages your workflows. Over weeks and months, it accumulates thousands of daily log entries. Without Auto-Dream, that knowledge sits in raw files — unsorted, unscored, disconnected. Your agent has the data but can't think about it.

Auto-Dream transforms your agent from a tool that executes commands into a partner that understands context, learns from history, and connects the dots you didn't see.

Why This Is Different

Architecture

┌──────────────────────────────────────────────────────────────────┐
│                                                                  │
│   ┌─────────────┐   ┌──────────────┐   ┌───────────────────┐   │
│   │   COLLECT    │──▶│ CONSOLIDATE  │──▶│     EVALUATE      │   │
│   │             │   │              │   │                   │   │
│   │ Scan 7 days │   │ Route layers │   │ Score importance  │   │
│   │ Flag markers│   │ Semantic dedup│   │ Forgetting curve  │   │
│   │ Extract     │   │ Assign IDs   │   │ Health metrics    │   │
│   │  insights   │   │ Link relations│  │ Generate insights │   │
│   └─────────────┘   └──────────────┘   └───────────────────┘   │
│                                                                  │
│                      ☽ Dream Cycle ☾                             │
└──────────────────────────────────────────────────────────────────┘
                              │
              ┌───────────────┼───────────────┐
              ▼               ▼               ▼
     ┌──────────────┐ ┌─────────────┐ ┌──────────────┐
     │  📊 Dashboard │ │ 🔔 Notify   │ │ 📝 Dream Log │
     │  HTML + Charts│ │ Push to chat│ │ Append report│
     └──────────────┘ └─────────────┘ └──────────────┘

Five Memory Layers

Features

🧠 Cognitive Dream Cycle

Runs automatically via cron (default: 4 AM daily). Three phases:

1. Collect — Scans unconsolidated daily logs (last 7 days), detects priority markers (⚠️ PERMANENT, 🔥 HIGH, 📌 PIN, <!-- important -->), extracts decisions / people / facts / projects / lessons / procedures / open threads

2. Consolidate — Routes each insight to the correct memory layer, performs semantic deduplication, assigns unique IDs (mem_NNN), creates relation links between connected entries

3. Evaluate — Scores importance using base_weight × recency × reference_boost / 8.0, applies forgetting curves (>90 days + low importance → archived, never deleted), calculates 5-metric health score, generates 1–3 non-obvious insights, writes dream report, sends notification

📊 Importance Scoring

Every entry gets a score on every dream cycle:

importance = (base_weight × recency_factor × reference_boost) / 8.0

• Recency: max(0.1, 1.0 - days/180) — gradual decay over 6 months
• References: log₂(count + 1) — logarithmic boost for frequently-referenced entries
• Markers: 🔥 HIGH doubles base weight; ⚠️ PERMANENT always scores 1.0

📉 Intelligent Forgetting

Memories aren't deleted — they're gracefully archived:

• Entry must be >90 days unreferenced AND importance < 0.3
• Compressed to a one-line summary in memory/archive.md
• Original ID preserved for relation tracking
• ⚠️ PERMANENT and 📌 PIN entries are immune

🕸️ Knowledge Graph

Entries are linked by semantic relations. The reachability metric measures graph connectivity:

• Union-find algorithm across all entry relations
• Detects isolated knowledge clusters
• Suggests cross-references to improve coherence

🩺 Health Score (5 Metrics)

health = (freshness×0.25 + coverage×0.25 + coherence×0.2 + efficiency×0.15 + reachability×0.15) × 100

🔔 Push Notifications

Dream results delivered to your chat automatically:

📊 Interactive Dashboard

A zero-dependency HTML dashboard with:

• Animated health gauge (0–100)
• 5 metric cards with color-coded status
• Memory distribution donut chart
• Importance histogram
• Health trend line chart (last 30 cycles)
• Force-directed knowledge graph visualization
• Recent changes, insights, suggestions, stale entries

Generate it: "Show memory dashboard"

🔄 Cross-Instance Migration

Move memories between MyClaw instances:

"Export memory bundle"    →  memory/export-20