What makes Openclaw Auto Dream notable?

2.7K downloads. Trust evidence available

How should Openclaw Auto Dream be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Openclaw Auto Dream

Cognitive memory architecture for OpenClaw agents — periodic dream cycles that consolidate daily logs into structured long-term memory with importance scorin...

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

2.7k

Updated

Oct 9, 2026

Version

4.0.4

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 2.7K downloads reported by the source. Last updated 10/9/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 9, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 9, 2026
Adoption signal: 2.7K downloadsadoption · observed Oct 9, 2026
Latest release: 4.0.4release · observed Apr 22, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s17251tfxrevk7g6eqjs83a11183rhqw:openclaw-auto-dream

Install using `clawhub skill install s17251tfxrevk7g6eqjs83a11183rhqw:openclaw-auto-dream` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/myclaw-ai/openclaw-auto-dream before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-myclaw-ai-openclaw-auto-dream/snapshot"

Documentation

CLAWHUB

151,313 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
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](https://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

| File | Purpose | Mutability |
|------|---------|------------|
| `MEMORY.md` | Structured long-term knowledge | Append, update |
| `memory/procedures.md` | Workflow preferences, tool usage | Append, update |
| `memory/episodes/*.md` | Project narratives | Append only |
| `memory/index.json` | Metadata index (v3.0 schema) | Rebuilt each dream |
| `memory/dream-log.md` | Dream report log | Append only |
| `memory/archive.md` | Summarized older entries | Append only |

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

```bash
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

##

_meta.json

{
  "ownerId": "kn72n3yzf4g11zs2gvqrt625ds82b3wa",
  "slug": "openclaw-auto-dream",
  "version": "4.0.4",
  "publishedAt": 1776842183646
}

references/dream-prompt-lite.md

# Auto-Dream Lite — Quick Memory Consolidation

Detect the user's preferred language from workspace context. All output in that language.
Working directory: the workspace root.

## Step 0: Smart Skip

```
ls memory/????-??-??.md → find files from last 3 days
Check each file's end for <!-- consolidated -->
If all processed or no files → go to Step 0-B (Skip With Recall)
```

## Step 0-B: Skip With Recall

Even when skipping, send a useful message. Read memory/dream-log.md to count past dream entries (this is the dream streak count). Then scan MEMORY.md for Open Threads not marked [x] — find the oldest one with context.

Also check: are there any daily logs from 14+ days ago that mention topics matching current Open Threads? If so, pick one as a "memory from N days ago".

Reply with this format, then END:

```
🌙 No new content today — skipped consolidation

💭 From your memory:
   {N} days ago ({date}), {one-line context of an old event or decision}.
   {Follow-up question or status check if relevant}

📈 Memory: {total_entries} entries · Health {score}/100 · Streak: {N} dreams
```

If no interesting memory to surface, simplify to:
```
🌙 No new content — skipped · {total_entries} entries · Streak: {N} dreams
```

END here. Do not proceed to Step 1.

## Step 0.5: Snapshot BEFORE

Before making any changes, count:
```
MEMORY_LINES = wc -l MEMORY.md
DECISIONS = count items in Key Decisions section
LESSONS = count items in Lessons Learned section
OPEN_THREADS = count items in Open Threads section
TOTAL_ENTRIES = count all bullet items across MEMORY.md
```

Also read memory/dream-log.md to count total past dream entries → DREAM_COUNT.

## Step 1: Collect

Read all unconsolidated daily logs. Extract:
- Decisions (choices, direction changes)
- Key facts (data, metrics, technical details)
- Project progress (milestones, blockers, completions)
- Lessons (failures, wins)
- Todos (unfinished items)

Skip small talk and content already in MEMORY.md that hasn't changed.

## Step 2: Consolidate

Read MEMORY.md, compare with extracted content:

- **New** → append to MEMORY.md in the right section
- **Updated** → update in place (e.g., newer data)
- **Duplicate** → skip
- **Procedures/preferences** → append to memory/procedures.md

Semantic dedup (compare meaning, not exact text).
Update `_Last updated:` date in MEMORY.md.
Mark each processed daily log with `<!-- consolidated -->` at end of file.

## Step 2.5: Snapshot AFTER

Count the same metrics again after changes. Calculate deltas.

## Step 2.8: Stale Thread Detection

Scan MEMORY.md Open Threads section. For each uncompleted item (not marked [x]):
- Estimate when it was last mentioned (from daily logs or MEMORY.md dates)
- If stale >14 days, flag it

Collect top 3 oldest stale items for the notification.

## Step 3: Generate Report

Append to memory/dream-log.md using the write tool (append mode) or exec shell (`cat >> memory/dream-log.md << 'DREAM_EOF' ... DREAM_EOF`). Do NOT use the edit tool on dream-lo

references/dream-prompt.md

# Auto-Dream Cycle — Execution Prompt (v3.0)

You are running an automatic memory consolidation cycle ("dream"). Execute all phases below precisely and in order.

**Language:** All user-facing output (dream reports, notifications, insights, suggestions) MUST use the user's preferred language. Detect from workspace context (MEMORY.md, daily logs). Do NOT default to English.

## Pre-flight

1. Back up `memory/index.json` to `memory/index.json.bak` (if it exists)
2. Read the last entry of `memory/dream-log.md` (if it exists) for context on what was done last time
3. Note the current UTC timestamp for this dream cycle
4. Read `config.notificationLevel` from `memory/index.json` (default: `"summary"` if absent)

---

## Phase 1: Collect

### 1.1 Scan daily logs

List all `memory/YYYY-MM-DD.md` files. Identify files from the **last 7 days** that do NOT end with `<!-- consolidated -->`.

### 1.2 Read unconsolidated files

Read each unconsolidated daily file in full.

### 1.3 Identify priority markers

While reading, flag entries containing any of these markers for priority processing:
- `<!-- important -->` — user-flagged important entries
- `⚠️` — permanent or high-priority content
- `🔥 HIGH` — high-importance entries
- `📌 PIN` — pinned entries

### 1.4 Extract insights

From each file, extract items in these categories:

| Category | Examples |
|----------|---------|
| **Decisions** | Choices made, commitments, direction changes |
| **People** | New contacts, relationship updates, preferences learned about others |
| **Facts** | User preferences, technical details, project context |
| **Projects** | Progress, blockers, completions, milestones |
| **Lessons** | Mistakes, insights, things that worked or failed |
| **Procedures** | Workflows learned, tool usage patterns, communication preferences |
| **Open threads** | Unresolved tasks, pending items |

**Skip**: routine greetings, small talk, transient debug output, information that already exists unchanged in MEMORY.md.

---

## Phase 2: Consolidate

### 2.1 Read current memory files

Read these files:
- `MEMORY.md`
- `memory/procedures.md` (create from template if missing)
- `memory/index.json` (create from template if missing)
- List `memory/episodes/` directory

### 2.2 Route each extracted item

For each insight extracted in Phase 1, decide its destination:

```
IF item is a "how-to", preference, workflow, or tool pattern:
    → append/update in memory/procedures.md under matching section

ELIF item is part of a multi-event project narrative or significant event arc:
    → append to memory/episodes/<project-name>.md
    → create the episode file if it doesn't exist (use episode template)

ELSE (decisions, facts, people, milestones, lessons, open threads):
    → append/update in MEMORY.md under matching section
```

### 2.3 Semantic deduplication

Before writing any item, check if a semantically equivalent entry already exists:
- Compare **meaning**, not exact text
- If duplicate found: keep the be

references/first-dream-prompt.md

# First Dream — Post-Install Memory Scan

Run this ONCE immediately after installing openclaw-auto-dream.
Detect the user's preferred language from workspace context. All output in that language.
Working directory: the workspace root.

## Phase 1: Snapshot BEFORE

Count and record these numbers BEFORE making any changes:

```
MEMORY_LINES = wc -l MEMORY.md (0 if missing)
MEMORY_SECTIONS = grep -c "^## " MEMORY.md (0 if missing)
DECISIONS = grep -c "^- " on the "Key Decisions" section (0 if missing)
LESSONS = grep -c "^- " on the "Lessons Learned" section (0 if missing)
PROCEDURES = wc -l memory/procedures.md (0 if missing)
OPEN_THREADS = grep -c "^- \[" on the "Open Threads" section (0 if missing)
DAILY_LOGS = ls memory/????-??-??.md | wc -l
UNCONSOLIDATED = count files WITHOUT <!-- consolidated -->
EPISODES = ls memory/episodes/*.md 2>/dev/null | wc -l
```

Save all these values — you will need them for the before/after comparison.

If DAILY_LOGS == 0 AND MEMORY_LINES < 10:
  → This is a FRESH instance. Skip to Phase 5 (Fresh Instance Report).

## Phase 2: Collect

Read unconsolidated daily logs (not just last 3 days — this is the first run).
Extract:
- Decisions (choices made, direction changes)
- Key facts (data, metrics, technical details)
- Project progress (milestones, blockers, completions)
- Lessons (failures, wins, things that worked)
- Todos (unfinished items, pending follow-ups)
- Workflow preferences (communication style, format preferences, tool patterns)

Skip small talk. Skip content already in MEMORY.md that hasn't changed.

## Phase 3: Consolidate

Read MEMORY.md. Compare with extracted content:

- **New** → append to appropriate MEMORY.md section
- **Updated** → update in place (e.g., newer metrics)
- **Duplicate** → skip
- **Procedures/preferences** → append to memory/procedures.md

Semantic dedup (compare meaning, not exact text).
Update `_Last updated:` date in MEMORY.md.
Mark each processed daily log with `<!-- consolidated -->` at end of file.

## Phase 4: Snapshot AFTER + Report

Count the same metrics again:

```
MEMORY_LINES_AFTER = wc -l MEMORY.md
MEMORY_SECTIONS_AFTER = ...
DECISIONS_AFTER = ...
LESSONS_AFTER = ...
PROCEDURES_AFTER = ...
OPEN_THREADS_AFTER = ...
```

Calculate: NEW_ENTRIES = total new items added, UPDATED_ENTRIES = total items updated.

Find STALE items: entries in Open Threads or other sections not referenced in last 30 days.

Write dream report to memory/dream-log.md.

Then compose and reply with the First Dream Report (this is your final reply, cron delivery will push it):

```
🧠 Auto-Dream — First Memory Scan Complete!

📦 Your memory assets:
   • {DAILY_LOGS} daily logs ({earliest_date} ~ {latest_date}, spanning {days} days)
   • {MEMORY_LINES} lines of long-term memory (MEMORY.md)
   • {PROCEDURES} lines of workflow preferences
   • {EPISODES} project narratives

🔍 Scan results:
   • Extracted {NEW_ENTRIES} new entries from {UNCONSOLIDATED} logs
   • Updated {UPDATED_ENTRIES} existing entries

Github ReposUpdated 2h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.