1. Executive Overview & 4W1H Architecture
This document presents the architectural framework and operational SOP for adopting Lola (lola-ai) within the Deep State of Mind (DSOM) AI orchestration system.
4W1H Framework Breakdown
-
WHAT (Definition & Analogy):
Lola (
lola-ai) is an open-source declarative package manager and synchronization engine designed specifically for AI agent skills and context packages. If an agent skill is an RPM package, Lola is the DNF package manager. It allows skills to be packaged as portable, version-controlled, dual-compliant filesystem bundles anchored bySKILL.md. - WHY (Strategic Rationale & Sovereignty): To enable seamless portability, versioning, and distribution of agent capabilities across heterogeneous execution environments (e.g., local control nodes, telecommunication bastions, enterprise clouds) without introducing hard dependencies on external SaaS registries, internet access, or WAN endpoints. This preserves 100% operational sovereignty and air-gapped readiness under DSOM Rule 16 and Rule 27.
- WHO (Target Audience & Governance): System Architects, GitOps Engineers, Ansible Automation Engineers, and AI Agents (e.g., Google Jules, Google Antigravity) operating under the DSOM Protocol.
-
WHERE (Location & File Anchors):
Declared globally at the repository root via
.lola-reqand locally vendored under.agents/skills/<skill-name>/. Automated deployment is orchestrated viaplaybooks/install.yml. -
HOW (Implementation & Synchronization SOP):
- Packaging: Skills are organised into standardised directory structures (
SKILL.md,scripts/,references/,tests/) with dual-compliant metadata (Lola + OKF v0.2). - Script Hermeticity: Python helper scripts embed PEP 723 inline script metadata (
# /// script ...) for self-containeduv runexecution. - Requirements Declaration:
.lola-reqenumerates all in-repo skills as locally vendored packages with optional sovereign Git remote fallbacks. - Idempotent Air-Gapped Sync: Ansible (
playbooks/install.yml) and shell scripts check forlolaCLI presence (which lola) before executinglola sync. If absent or offline, sync is skipped gracefully, falling back to local vendored skills.
- Packaging: Skills are organised into standardised directory structures (
2. Dual-Compliant Metadata Standard (Lola + OKF v0.2)
Every skill package in.agents/skills/<skill-name>/SKILL.md implements a unified frontmatter combining Lola package manager attributes with OKF v0.2 trust pillars:
3. Standard Skill Package Directory Layout
All skills adhere to the following directory structure:4. Declarative Repository Manifest (.lola-req)
The root .lola-req manifest configures local vendored skill resolution with one module reference per line (lola-ai>=0.1.0 CLI format):
5. Air-Gapped Idempotent Automation Snippets
Ansible Playbook (playbooks/install.yml)
POSIX Shell Script Snippet
6. Complete Skill Package Summary Matrix (46 Skills)
Deep State of Mind (DSOM) For My AI Protocol | Harisfazillah Jamel (LinuxMalaysia) | 2026-09-06 Standard: UK English | DBP-standard Bahasa Melayu Malaysia (Piawai) | GNU General Public License v3.0