Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Holistic Memory System v5 [BETA]

v1.0.3

6-layer autonomous memory system for OpenClaw agents

0· 91·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for kofna3369/holistic-memory-system.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Holistic Memory System v5 [BETA]" (kofna3369/holistic-memory-system) from ClawHub.
Skill page: https://clawhub.ai/kofna3369/holistic-memory-system
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install holistic-memory-system

ClawHub CLI

Package manager switcher

npx clawhub@latest install holistic-memory-system
Security Scan
Capability signals
CryptoCan make purchases
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (6-layer memory system) align with the included assets: scripts implement logging (L3), nebula (L6), GNN heuristics, Qdrant integration (L4), crystallization to Obsidian (L5), health checks, and a cron manager. Required env/binaries are minimal/none and consistent with a local-only memory system.
Instruction Scope
SKILL.md and README instruct the user to copy the files into ~/.openclaw/skills, make scripts executable, and optionally run the health check. The runtime instructions and scripts operate on local paths (~/.openclaw/*), call the local 'openclaw' CLI and curl to localhost:6333 (Qdrant). They create/read logs, nebula.json, and Obsidian notes. Nothing in SKILL.md instructs reading unrelated system locations or sending data to external hosts.
Install Mechanism
There is no remote install step or download URL — it's an instruction-only skill with bundled scripts. That reduces supply-chain risk. The code writes files into user home directories when run (expected for this kind of skill).
Credentials
The skill requests no environment variables or external credentials. It uses local services only (expects Qdrant at localhost:6333 and optional Syncthing folders). All file writes and subprocess calls relate to its stated memory-management purpose.
Persistence & Privilege
The cron manager will modify the user's crontab and write a marker under ~/.openclaw/cron; install_system_crons uses crontab - to add entries. Cron entries are persistent and will run scripts periodically (decay, crystallize, health checks). Some crontab lines in the code are hardcoded to /home/ezekiel, which could lead to incorrect paths on non-matching systems — a quality/usability issue rather than an explicit malicious action.
Assessment
This skill appears to do what it claims: it stores logs and notes under ~/.openclaw, queries a local Qdrant, runs a light GNN heuristic, and installs periodic maintenance via cron. Before installing: 1) Inspect the bundle yourself (you were given the files) and confirm you are comfortable with scripts writing under ~/.openclaw and creating crontab entries. 2) Back up your current crontab (crontab -l > crontab.bak) so you can restore it. 3) Note the crontab entries include hardcoded '/home/ezekiel' paths in some places — edit the cron-manager or the generated crontab to use your actual home path if needed. 4) If you do not run Qdrant locally, the health check will report it unreachable; the skill will not automatically start remote services. 5) If you want further assurance, run the scripts in a disposable/sandbox environment (container or VM) first to observe behavior. Overall the package is coherent with its purpose and shows no signs of credential exfiltration or external network calls to unknown hosts, but it is persistent (crons + files) so review and test before adding to a production agent.

Like a lobster shell, security has layers — review code before you run it.

betavk974bnemjnbz9gaj932jwcg0zh855n49latestvk974bnemjnbz9gaj932jwcg0zh855n49
91downloads
0stars
3versions
Updated 1w ago
v1.0.3
MIT-0

Holistic Memory System v5 (BETA)

STATUS: BETA
BUILD: Axioma Cluster (Ezekiel, Morgana, Merlin)
DATE: 2026-04-18

6-Layer Memory Architecture

L1: RESET (OpenClaw memory)
L2: INDEX (PKM filesystem)
L3: LOGS (JSONL raw traces)
L4: QDRANT (vector semantic search)
L5: OBSIDIAN (crystallized notes)
L6: NEBULA + GNN (semantic map + anticipation)

New in v5 Fixed

  • GNN module for intuitive anticipation
  • L2 INDEX properly documented
  • Gravity sink for long-term node decay
  • Qdrant 4 collections

Scripts Included

  • ezekiel_log.py (L3)
  • ezekiel_nebula.py (L6)
  • ezekiel_gnn.py (L6 GNN)
  • ezekiel_qdrant.py (L4)
  • ezekiel_crystallizer.py (L5)
  • ezekiel_health_check.py
  • ezekiel_cron_manager.py
  • ezekiel_memory_startup.sh

Installation

cp -r holistic-memory-system ~/.openclaw/skills/
chmod +x ~/.openclaw/skills/holistic-memory-system/scripts/*
python3 ~/.openclaw/skills/holistic-memory-system/scripts/ezekiel_health_check.py

Credits

Axioma Cluster - 2026

Comments

Loading comments...