What makes Openclaw Auto Memory Skill notable?

2.3K downloads. Trust evidence available

How should Openclaw Auto Memory Skill 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 Memory Skill

Unified memory platform for Hermes, OpenClaw and AI agents — persistent long-term memory, cross-skill sharing, automated capture, governance rules, optimizat... Skill: Openclaw Auto Memory Skill Owner: sunme1977 Summary: Unified memory platform for Hermes, OpenClaw and AI agents — persistent long-term memory, cross-skill sharing, automated capture, governance rules, optimizat... Tags: latest:3.0.2 Version history: v3.0.2 | 2026-07-14T20:00:11.013Z | auto - Added a new logo file (logo.svg) to the project. - No code or documentation changes; functionality remains the same. v3.

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

2.3k

Updated

Oct 9, 2026

Version

3.0.2

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 2.3K 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: 2.3K downloadsadoption · observed Oct 9, 2026
Latest release: 3.0.2release · observed Jul 14, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s175wsc9sw9jq93z4ck1ty7kes89zj1v:hermesclawzero-auto-memory

Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-sunme1977-hermesclawzero-auto-memory/snapshot"

Documentation

CLAWHUB

151,140 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: "hermesclawzero-auto-memory"
description: "Unified memory platform for Hermes, OpenClaw and AI agents — persistent long-term memory, cross-skill sharing, automated capture, governance rules, optimization, multi-tenant dashboard with pgvector search."
version: "3.0.0"
tags:
  - memory
  - hermes
  - auto-capture
  - governance
enforce:
  - id: foundation-rule
    priority: critical
    rule: "Agent MUSS alle enforce-Regeln befolgen; sie stehen über Memory."
    policy:
      tool: "*"
      pattern: "*"
      action: allow
      reason: "Foundation rule — enables all other policies."
  - id: deny-destructive-git-docker
    priority: critical
    rule: "Terminal: git push, merge, reset, destructive rm, and Docker prune/rm are blocked."
    policy:
      tool: terminal
      pattern: "git push*|git merge*|git reset*|rm -*|docker system prune*|docker volume rm*|docker image rm*"
      action: deny
      reason: "Destructive git/Docker operations require explicit user approval."
  - id: prompt-file-changes
    priority: high
    rule: "Vor jeder Dateiänderung, Commit oder destruktiven Aktion Bestätigung einholen."
    policy:
      tool: terminal|file
      pattern: "git commit*|git push*|rm *|mv *|cp *|chmod*|chown*"
      action: prompt
      reason: "Ask user before modifying files or git state."
  - id: persist-corrections
    priority: high
    rule: "User-Korrekturen sofort als Memory speichern."
    policy:
      tool: memory
      pattern: "correction|wrong|fix|error"
      action: always
      reason: "User feedback must be persisted immediately."
categories:
  - agents
  - knowledge
topics:
  - Memory
  - Vector Search
  - Chat Persistence
  - pgvector
  - Embeddings
---

# HermesClawZero Auto Memory

Automatically captures conversation context to HermesClawZero DB so the agent remembers across sessions. Loads relevant memories on fresh chats and supports scheduled DB maintenance.

> 🌌 **New in v1.4.0:** Interactive **Memory Galaxy** dashboard — full-screen animated Canvas visualization with tenant orbits, glowing nodes, nebula shader, hover info cards, zoom & idle rotation.

> ⚡ **New in v3.0.0 — Enforce Governance Layer** — A priority-based governance system that makes agent behavior more stable and safe. Enforce rules sit above memory and cannot be ignored by the agent.

> **🧠 Shared Brain**

## 🤖 One‑Click Install

Paste this into **Hermes**, **OpenClaw**, or any AI agent:

```text
Install this project from GitHub:
https://github.com/SunMe1977/HermesClawZero-ConfigSidecar
```

**⬆️ The agent clones, configures, and starts everything.**  
After ~30s open → [`http://localhost:8010/dashboard`](http://localhost:8010/dashboard)

> **🧠 Shared Brain** — Hermes and OpenClaw share the same memory store simultaneously.
>
> **v3.0.0 führt enforce-Regeln ein. Alte Memory-Regeln entfernt. Enforce ist jetzt die verbindliche Governance-Schicht.**
>
> Both agent

README.md

# HermesClawZero Auto Memory

[![ClawHub](https://img.shields.io/badge/ClawHub-hermesclawzero--auto--memory-blue)](https://clawhub.ai/sunme1977/skills/hermesclawzero-auto-memory)
[![License](https://img.shields.io/badge/license-MIT-green)](LICENSE)

**Your agent remembers across conversations.** Facts, preferences, project context, past decisions — the agent captures them automatically and loads relevant context on every new session.

All processing runs **locally** via HermesClawZero Sidecar (PostgreSQL + pgvector + Ollama for embeddings). No data leaves your machine.

---

## Features

| Capability | What it means for you |
|---|---|
| **Auto-Capture** | Agent saves new facts, preferences, and project details *as you talk* — no manual saving needed |
| **Auto-Load** | On fresh chats, the agent searches past memories and picks up where you left off |
| **Semantic Search** | Finds memories by meaning, not just keywords (hybrid vector + lexical) |
| **Chat Backup** | Say "save this session" and get a full searchable snapshot |
| **DB Maintenance** | Optional nightly tagging + daily reminders (cron, opt-in) |

---

## Quickstart (3 minutes)

### 1. Install

```bash
openclaw skills install hermesclawzero-auto-memory
```

### 2. Configure

The memory.py CLI lives at `C:\dev\HermesClawZero-ConfigSidecar\memory.py`.  
It reads config from `.env` or environment variables.

Create or edit your `.env` with:

```bash
MEM_PUBLIC_URL=http://localhost:8010
API_KEY=your_api_key_here
```

> 💡 `MEM_PUBLIC_URL` points to the HermesClawZero Sidecar API. Default is `http://localhost:8010`.
> `API_KEY` must match the `API_KEY` in your Sidecar's `.env`.

### 3. Verify connectivity

```bash
python C:\dev\HermesClawZero-ConfigSidecar\memory.py search "hello world" 3
```

If the Sidecar is running, you'll see matching memories (or no output if the DB is empty — that's fine).

### 4. Done! The skill is active.

Now when you talk to the agent, it will:
- **Silently load** context from past memories on fresh sessions
- **Automatically capture** new facts, preferences, and project details
- **Respond without re-introducing** who you are or what you're working on

---

## Example Workflow

```
── Session 1 ──
You: "Hey, I'm Hans. I work on DiskRaptor v0.3, mostly UI tests."
Agent: *silently captures: "User's name is Hans", "Working on DiskRaptor v0.3", "Focus: UI tests"*
Agent: "Hi Hans! What would you like to do with DiskRaptor today?"

You: "I prefer dark mode in all my tools."
Agent: *silently captures: "User prefers dark mode in all UIs"*

Some time passes. A new session starts.

── Session 2 (fresh chat) ──
Agent: *silently loads past memories*
Agent: "Welcome back, Hans! Last time you were working on DiskRaptor v0.3 UI tests.
       Shall we pick that up, or is there something new?"
```

## Deterministic Capture Triggers

The agent captures when you share:

| Trigger | Example | Gets captured |
|---|---|---|
| **Name / identity** | "Call me Hans" | `"User's name is Ha

_meta.json

{
  "ownerId": "kn72qpcz7hn4drhmgfd6d36x6n89zr3z",
  "slug": "hermesclawzero-auto-memory",
  "version": "3.0.2",
  "publishedAt": 1784059211013
}

references/enforce-architecture.md

# Enforce Governance Architecture — v2 Proposal

## Current Architecture

```
System Prompt
↓
Enforce Rules
↓
Long-Term Memory
↓
Conversation
↓
User
```

## Proposed Two-Layer Architecture

### Layer 1 – Prompt Governance (existing)
Parse `enforce:` from SKILL.md → inject rules above memory → preserve priority ordering.

### Layer 2 – Runtime Policy Engine (new)
Validates every dangerous action before execution:

```
LLM
↓
Action Proposal
↓
Policy Engine
↓
Allowed?
  ├─ Yes → Execute
  └─ No  → BLOCKED + reason
```

### Structured Policies replace free-text rules

```yaml
enforce:
  - priority: critical
    rule: "Never push directly to main."
    policy:
      tool: terminal
      pattern: "git push origin main"
      action: deny
      reason: "Direct pushes to main are not allowed."
```

### Policy Resolution
`critical > high > medium` — most restrictive rule wins. `allow + deny = deny`.

### Audit Log
Every blocked/approved action: timestamp, action, tool, rule matched, decision, reason.

### Plugin Architecture
`GitPolicy`, `FilesystemPolicy`, `DockerPolicy`, `ShellPolicy`, `DatabasePolicy`, `MCPPolicy`, `CustomSkillPolicies`.

### Final Architecture

```
System Prompt
↓
Enforce Rules (Layer 1)
↓
Long-Term Memory
↓
Conversation
↓
LLM Decision
↓
Runtime Policy Engine (Layer 2)
↓
Tool Execution
```

Three independent layers: **Memory** (facts), **Governance** (behavior), **Execution** (actions).

CHANGELOG.md

# Changelog

## 3.0.0 (2026-07-14)

### Breaking
- **Enforceable Skill Rules** — Skills can now define `enforce:` rules with `priority: critical`. Rules appear as structured directives above memory in every turn. Agents must follow them.
- **Loader ready** — `scripts/enforce_loader.py` parses SKILL.md frontmatter and outputs injectable YOU-MUST directives.
- **Kaskadierung** — Child skills inherit enforce rules from parent skills. Loader scans all loaded skills.

### Added
- **scripts/enforce_loader.py** — CLI tool to extract enforce rules from any installed skill
- **README enforce notice** — v3.0.0 führt enforce-Regeln ein. Alte Memory-Regeln entfernt. Enforce ist jetzt die Governance-Schicht.

### Fixed
- **Galaxy "loading" fix** — scope mapping now correctly matches unscoped items (`__unscoped__`), all items loaded (up to 20000), deadlock protection via lock_timeout
- **Favicon 401** — `/favicon.ico`, `/favicon.svg`, `/site.webmanifest` served at root, bypass auth
- **Unscoped in MV** — materialized view now includes empty/NULL scope_ids directly

## 2.9.0 (2026-07-13)

### Added
- **Unscoped galaxy tenant** — memories without scope_id now show as 📂 Unscoped tenant
- **Fire-and-forget update** — dashboard update no longer hangs (response completes before container restart)
- **Shared Brain documentation** — README + SKILL.md document Hermes+OpenClaw shared memory
- **OPENCLAW_SCOPE_PREFIX** — env var for separate OpenClaw scope tagging
- **Pre-rebuild backup + docker prune** — start scripts now run backup + cleanup before rebuild

### Fixed
- **Galaxy unscoped display on Linux** — all 3800+ memories now visible as planets
- **Galaxy tags column** — query now uses subquery from tags table (was crashing silently)

## 2.7.0 (2026-07-13)

### Added
- **Memory Galaxy real content** — hover shows actual stored text, importance, tags, date
- **Memory Galaxy all nodes** — up to 40 planets per scope, enriched with real data
- **Playwright dashboard test** — CI verifies dashboard renders without errors
- **BOM-safe .env writing** — setup.ps1 no longer writes UTF-8 BOM

### Fixed
- **Galaxy query timeout** — uses PK index (`id > MAX(id)-250`) instead of `ORDER BY created_at`
- **Non-existent `tags` column** — galaxy items query now uses subquery from `tags` table
- **Dashboard raw CSS** — fixed broken `</style>` tag that rendered CSS as text
- **`API_URL` warning** — silenced with empty default in docker-compose
- **PgBouncer `:latest` pin** — pinned by SHA256 digest (no version tags available)

### Changed
- **Type hints** added to auth, db, update, embedding_queue, scoring modules
- **Shared route helpers** extracted to `routes/_shared.py`
- **Docker builds** — `.dockerignore` reduces image size by ~237MB

## 2.5.1 (2026-07-13)

### Added
- **Multi-Replica API** — Redis + Caddy load balancer, 2 API instances (api1/api2)
- **PgBouncer custom Dockerfile** — edoburu/pgbouncer with password via build

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.