linuxmalaysia/deep-state-of-mind

v10.4.0GPL-3.0-or-later

Deep State of Mind (DSOM) Sovereign AI Agent Cognitive Architecture, Open Knowledge Format (OKF), and Model Context Protocol (MCP) toolset.

GitHub Open Website

okf_version: 0.2 type: documentation title: "📜 Changelog: Deep State of Mind (DSOM) For My AI" timestamp: "2026-08-20T23:25:00Z" topics: ["dsom", "documentation", "changelog", "okf"] spec_version: "0.2" description: "OKF-compliant documentation for CHANGELOG.md." resource: "file:///CHANGELOG.md"

📜 Changelog: Deep State of Mind (DSOM) For My AI

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

Added

Start AI Agents Setup Prompt & Execution Protocol (Entry Point 23): Authored docs/START-AI-AGENTS-PROMPT.md providing a universal, copy-pasteable setup prompt and execution protocol for human operators to initialize any AI agent (Google Jules, Antigravity, Claude, Cursor, Copilot, Gemini) into full DSOM Protocol compliance for first-time project setup.
Mintlify User Manual Architecture & Downstream Sync Guide (Entry Point 22): Authored docs/governance/MINTLIFY-USER-MANUAL-SYNC-GUIDE.md detailing the 3-manual stratification model (User/Admin/Developer Manuals), compiler requirements, and the universal drop-in AI agent prompt for other repositories.
Mintlify One-Way Docs Sync Pipeline & Safety Guards (Rule 31 & Entry Point 21): Implemented scripts/sync_docs.py and .github/workflows/sync-docs.yml to safely publish docs-source/ to the downstream docs repo linuxmalaysia/my-knowledge-brain serving https://harisfazillah.mintlify.site.
5-Layer Safety Guardrails (A–E): Enforced source validation, minimum 5 .mdx files floor, navigation integrity checking every page path in docs.json, diff preview with a 10-deletion cap (ALLOW_LARGE_DELETIONS override), and dry-run preview mode.
Concise Output Standard (DTS 0.1): Appended DTS 0.1 output formatting blocks across Universal Gateways (.cursorrules, CLAUDE.md, .github/copilot-instructions.md, AGENTS.md, .agents/AGENTS.md).
Agent Plugins 1.0.0 Standard Specification (Entry Point 20): Authored docs/governance/DSOM-AGENT-PLUGINS-SPECIFICATION.md detailing universal packaging, runtime interoperability, and directory layout standard (plugin.json, mcp.json, skills/).
Standard Manifests: Added canonical root plugin.json (schema 1.0.0) and mcp.json mapping DSOM FastMCP Server with ${PLUGIN_ROOT} and ${PLUGIN_DATA} environment variable expansion.
Agent Plugin Packager Skill: Created .agents/skills/agent-plugin-packager/SKILL.md (OKF v0.2) to automate packaging of skills and MCP tools into portable plugins.
Rule 30 & Rule 31 Codification: Codified Rule 30 (Agent Plugins 1.0.0 Specification & Packaging Mandate) and Rule 31 (Mintlify One-Way Docs Sync & Safety Guards Mandate) in .agents/AGENTS.md and root AGENTS.md.
Automated Conformance Tests: Implemented tests/test_agent_plugins_spec.py and tests/test_sync_docs.py ensuring 100% compliance with Agent Plugins 1.0.0 and Mintlify Safety Guards.

[10.4.0] - 2026-08-22

Added

AI Guardrails Master Reference (Entry Point 18): Authored docs/governance/AI-GUARDRAILS-MASTER-GUIDE.md detailing defense-in-depth safety, threat vectors, Tri-Phasic interception, and dual-path execution.
Guardrail Catalog & Submission Review: Authored docs/governance/DSOM-GUARDRAILS-CATALOG-SUBMISSION-REVIEW.md cataloging 10 sovereign guardrails for human review and PyPI publishing.
Downstream DSOM Protocol Compliance Mandate: Authored docs/governance/DOWNSTREAM-DSOM-COMPLIANCE-MANDATE.md establishing the 5-Pillar Sovereign Anchor Architecture for downstream repositories.
Standalone guardrails-ai-dsom Package: Implemented, tested, and packaged 10 custom validators under PEP 517/621 in tools/guardrails-ai-dsom/.
Universal Gateway Matrix: Deployed root .cursorrules, enriched CLAUDE.md, upgraded .github/copilot-instructions.md, and aligned root AGENTS.md.
Git Pre-Commit Guardrails Hook: Implemented tools/install_git_guardrails.py running guardrails-ai-dsom on every Git commit.
Episodic Resume Protocol Entry Point 19: Registered docs/DSOM-EPISODIC-RECORD-TEMPLATE.md across START-HERE.md, SUMMARY.md, mkdocs.yml, and llms.txt.
Dual-Path Custom Validator Mandate (Rule 29): Codified Rule 29 in .agents/AGENTS.md and root AGENTS.md.
Downstream Cross-Agent Honor Mandate (Rule 28): Codified Rule 28 enforcing Universal Gateway Matrix and pre-commit hooks.
Opportunistic OKF v0.2 Migration Mandate (Rule 6): Codified the requirement to migrate documents to OKF v0.2 (sources, generated, verified, status, stale_after) opportunistically during standard editing workflows, strictly forbidding token-wasting mass full-corpus rewrites.
Downstream Asymmetry & Cognitive State Preservation Mandate (Rule 28): Codified the architectural rule governing the Reference Forge (comprehensive baseline engine) vs. Minimal Downstream Adoption (lean 6-pillar footprint keeping client source code >90% of repository volume).
DSOM Cognitive State Preservation & Minimal Downstream Adoption Architecture: Authored foundational human-to-Gemini proposal (docs/governance/DSOM-COGNITIVE-STATE-PRESERVATION-PROPOSAL.md) establishing the Asymmetry Principle (comprehensive framework engine vs. lightweight, zero-bloat downstream project adoption), 6 core pillars of persistent memory, and Diátaxis 4-Quadrant navigation compass integration in START-HERE.md.
Dedicated Open Knowledge Format (OKF) Adoption Guide: Authored a comprehensive, human- and AI-readable specification document (docs/OKF-ADOPTION-GUIDE.md and references/OKF-ADOPTION-GUIDE.md) detailing Google Cloud's OKF v0.1 and v0.2 standards, token optimization, progressive disclosure, trust signals, and step-by-step adoption for DSOM repositories.
OKF Context Engine Integration: Elevated OKF as Entry Point 17 in START-HERE.md and integrated OKF documentation references across README.md, docs/README.md, SUMMARY.md, and mkdocs.yml.
Cross-Skill OKF Compliance Mandate: Updated key operational skills (okf-frontmatter-injector, dsom-policy-adopter, dsom-knowledge-ingester, dsom-bootstrap, dsom-project-cloner, openwiki-compiler, palace-auditor) to explicitly embed okf-frontmatter-injector execution (python tools/apply_okf_frontmatter.py) as a mandatory step whenever new Markdown documents or memory assets are created or modified.
Native Python OpenWiki Emulator (Rule 27): Built tools/openwiki_emulator.py with --init, --update, --search, and --export-graph CLI modes, eliminating Node.js runtime friction (version mismatches, native C++ compilation, UAC elevation hangs, and API rate limits). Reclaimed 135.3 MB of local disk space.
FastMCP OpenWiki Knowledge Bridge: Integrated OpenWiki resources (dsom://openwiki/skeleton, dsom://openwiki/quickstart), governance rulebook (dsom://governance/agents), and the search_openwiki(query) RAG search tool into tools/mcp/server.py.
OpenWiki Code Samples & Prompt Templates: Authored comprehensive code samples and reusable AI prompt templates in HOWTO-OPENWIKI.md, OPENWIKI-INTEGRATION-GUIDE.md, and HOWTO-MCP-SERVER.md.
Read the Docs Integration: Configured .readthedocs.yaml at the repository root using Ubuntu 24.04, Python 3.13, and MkDocs integration to automatically build the project's documentation on Read the Docs. Added comprehensive unit tests in tests/test_readthedocs_config.py to verify configuration integrity.
Executor Modularity & Windows WSL2 Bridge: Redefined the Third Pillar in GITOPS-AIOPS-ANSIBLE-STRATEGY.md and README.md from strictly "Ansible" to the modular "Executor". Formalized uv (Python), npm (Web), and pandoc (Docs) as valid executors. Formally established WSL2 as the mandatory local Control Node (Execution Bridge) for Windows-only projects to allow seamless execution of Linux tooling without a dedicated remote jumphost.
AI Initialization Sequence Document: Published docs/governance/AI-INITIALIZATION-SEQUENCE.md to explicitly document the mechanical boot process (AGENTS.md -> Brain -> START-HERE -> Governance -> Skills). Linked directly in README.md and GitBook as mandatory reading for both humans and AI.
Knowledge-First Protocol Injected into Core Governance: Embedded the protocol directly into docs/governance/AI-MASTER-PROTOCOL.md, DSOM-EFFICIENCY-PROTOCOLS.md, and OKF-MIND-OPTIMIZATION.md to ensure it operates as a fundamental foundational mechanic, not just an isolated SOP.
Knowledge-First Protocol Embedded as Entry Point: Elevated the Knowledge-First Discovery Protocol to the 11th Entry Point in START-HERE.md and explicitly mandated its inclusion in docs/HOWTO-CLONE-DSOM-PROJECT.md to ensure it operates as the baseline cognitive engine for all new DSOM clones.
Temporal Knowledge Verification (Rule 21): Established a mandate requiring AI agents to check OKF document timestamps. If deemed stale, the AI must cross-reference external sources, compare findings, and await human verification before executing or modifying knowledge.
Local Knowledge-First Protocol (Rule 20): Mandated that the AI must perform OKF frontmatter search via grep_search on .agents/brain/ and docs/ before issuing exploratory terminal commands, formalizing context preservation prior to node execution.
Token Efficiency & Byte-Capped Execution: Engineered the dsom-token-calculator AI skill and documented proofs showing DSOM Progressive Disclosure saves 96% of context tokens.
Android Termux Extrapolation: Benchmarked Android FUSE latency directly (tools/bench_brain.py) to map expected mobile pipelines.
Cognitive Flow Map: Injected a visual Mermaid mapping into ZERO-GLOBAL-MEMORY.md to diagram Sovereign AI memory retrieval.
9-Point System Synchronization: Expanded and formalized the project entry architecture from 4 to 9 entry points in START-HERE.md. Automatically cross-referenced all 9 entry point documents upstream and downstream.
Legacy Upgrade Formalization: Extracted legacy DSOM upgrade protocols into a dedicated OKF-compliant document (docs/HOWTO-UPGRADE-LEGACY-DSOM.md).
Dual Documentation Sync: Added Rule 14 to .agents/AGENTS.md explicitly requiring that new governance and architecture documents must be symmetrically mapped into both SUMMARY.md (GitBook) and mkdocs.yml (MkDocs) to prevent orphaned resources.
Release Automation: Encapsulated the DSOM release architecture into a new AI skill (.agents/skills/dsom-release-manager). This procedural SOP governs GitOps tagging, Changelog promotion, Ledger footer sync, and programmatic release deployment via the GitHub (gh) and GitLab (glab) CLI tools.
LLMs.txt Standard Adoption: Adopted the llmstxt.org specification by generating an llms.txt file at the repository root. This provides external LLMs and AI crawlers with a structured, machine-readable index of the Sovereign Engine's core entry points, governance ledgers, protocols, and multi-agent skills.
Gemini Gem Integration Guide: Authored docs/HOWTO-CREATE-DSOM-GEMINI-GEM.md providing step-by-step instructions on permanently anchoring the DSOM protocol within the Google Gemini interface via custom Gems.
Episodic Record Template: Established docs/DSOM-EPISODIC-RECORD-TEMPLATE.md to standardize the saving and restoring of AI Cognitive States between ephemeral chat sessions.
Episodic Resume Protocol: Added Rule 18 to .agents/AGENTS.md explicitly requiring the AI to generate a [DSOM EPISODIC RECORD] summary block at the end of complex workflows or before session termination to ensure strict context continuity.
START-HERE.md Enhancement: Added a comprehensive index of all mandatory DSOM operational skills under the Procedural Skill Entry Point to ensure complete porting when scaffolding new projects.
DSOM Token Efficiency Report: Executed a local token audit using uv and tiktoken, proving a 96% reduction in context window bloat. The results were permanently documented in docs/governance/DSOM-TOKEN-EFFICIENCY-REPORT.md.
Token Calculator Skill: Created .agents/skills/dsom-token-calculator to allow the AI to programmatically verify the token footprint of files or directories prior to ingestion, enforcing Rule 10 (Byte-Capped Executions).
Byte-Capped Execution Framework: Documented the technical layout and deployment model of the DSOM Token Calculator Skill in docs/governance/BYTE-CAPPED-EXECUTION-FRAMEWORK.md.

[10.4.0-governance] - 2026-07-12

Added

Governance Rule Update: Added the GitHub & GitLab CLI Authentication rule to .agents/AGENTS.md to formally guardrail multi-account API interactions and prevent context-blind CLI failures (e.g., linuxmalaysia vs songketmailsdnbhd).
Initialization Guide Update: Enhanced HOWTO-CLONE-DSOM-PROJECT.md to clarify the mandatory pre-requisite of git init and brain artifact creation (task.md, walkthrough.md) before running reanimation rituals.
Cognitive Architecture Update: Adopted the OKF Mind Optimization protocol (Progressive Disclosure, Artifact Pyramid, and Semantic Routing) and codified its procedural execution constraints into the Core Rulebook (.agents/AGENTS.md).
Generative Engine Optimization (GEO): Published docs/governance/GENERATIVE-ENGINE-OPTIMIZATION.md and mandated GEO formatting requirements (authoritative tone, statistical density, H2 queries) in .agents/AGENTS.md. Established the root llms.txt file as the standard AI ingestion sitemap.
Governance Policy: Published docs/governance/OKF-MIND-OPTIMIZATION.md to formally define the theoretical mechanics of the new cognitive architecture.
Skill Architecture Formalization: Created docs/governance/AI-SKILL-ARCHITECTURE.md to define semantic routing, progressive disclosure, and the project's active skills. Added Rule 12 to .agents/AGENTS.md enforcing skill usage as standard operational SOPs.
Architecture Deconstruction: Published docs/governance/DSOM-ARCHITECTURE-ANALYSIS.md outlining the repository's structural mapping, isolation namespaces, and operational loops. Enhanced the document with a detailed mapping of the 3 Core Operational Cycles (SOD, Runtime Loop, EOD).
Universal Sovereign Signature: Engineered the dsom-signature-injector skill to dynamically inject ownership and GPL v3.0 license signatures into all .md, .sh, .ps1, and .yml files based on their last modified timestamps.
Signature Mandate: Added Rule 13 to .agents/AGENTS.md enforcing that all readable files generated or modified by an AI must be processed by the dsom-signature-injector.
Automation Audit Ledger: Created docs/governance/AUTOMATION-AUDIT-LIST.md to catalog, index, and describe all 50+ .sh, .ps1, and .yml files in the repository for human security audits.
Python UV Environment Guide: Adopted and published docs/governance/PYTHON-UV-ENVIRONMENT-GUIDE.md to mandate the use of uv for managing isolated Python environments across the DSOM framework, avoiding global PATH conflicts.
NOSS Integration Guide: Published docs/governance/NOSS-INTEGRATION-GUIDE.md detailing the protocol for adopting National Occupational Skills Standards (NOSS) as compliance payloads within the DSOM architectural baseline. Includes token optimization strategies for handling original PDFs.
Project Entry Points: Authored and subsequently expanded START-HERE.md at the repository root to establish the 9 primary onboarding entry points (Engineering, Cognitive, Crawlers, Operations, Legacy Upgrades, Subagent Swarms, Procedural Skills, LLM WIKI, Defensive GitOps) for humans and AI adopting the DSOM framework.

Changed

Standard Developer Metadata: Upgraded the dsom-signature-injector skill to utilize industry-standard metadata headers for executable scripts and configuration files. Replaced flat comments with structured blocks for Bash/YAML (# =====) and native Comment-Based Help blocks for PowerShell (<# .SYNOPSIS #>), improving human readability and compatibility with native CLI tools like Get-Help.

Fixed

Signature Injector Shebang Conflict: Upgraded dsom-signature-injector to intelligently detect and preserve #! shebangs on Line 1 of executable scripts, injecting the signature on Line 2. Executed a mass-cleanup script to repair 18 affected Linux/Bash scripts (.sh).
PowerShell 5.1 Compatibility: Fixed syntax errors in Join-Path invocations inside tools/reanimate.ps1. The legacy PS 5.1 engine does not support more than two arguments natively for Join-Path. This patch allows seamless scaffolding into environments running native Windows PowerShell 5.1.

[10.3.0-bootstrap] - 2026-07-04

Added

DSOM Bootstrapping Skill: Adopted the dsom-bootstrap AI skill (.agents/skills/dsom-bootstrap) to enable the agent to natively deploy this framework's standards into external repositories.
DSOM Project Cloner Skill: Built the dsom-project-cloner AI skill to autonomously scaffold new DSOM projects from the baseline repository, automating the HOWTO-CLONE-DSOM-PROJECT.md blueprint.
Bootstrapping Blueprint: Added docs/HOWTO-DSOM-BASELINE.md as the definitive origin documentation for bootstrapping.
Project Cloning Guide: Added docs/HOWTO-CLONE-DSOM-PROJECT.md as the master blueprint for scaffolding a new project from this highly customized Sovereign Engine.
Diagnostic Tooling: Created standardized tools/diagnostic.ps1 and tools/diagnostic.sh for post-bootstrap verification.

Changed

Governance Restructuring: Relocated 13 core protocol documents (including AI-MASTER-PROTOCOL and DIGITAL-SOVEREIGNTY-MODEL) from docs/ to docs/governance/ to align with the bootstrapping framework.
Toolchain Standardization: Fixed legacy syntax bugs in pre-flight scripts (e.g., PS 5.1 Join-Path positional errors, unclosed quote issues) and updated paths to reflect the new docs/governance/ directory.
Registry Updates: Updated SUMMARY.md and mkdocs.yml to point to the relocated governance documents.

[10.2.0-okf] - 2026-06-19

Added

Open Knowledge Format (OKF) v0.1. Fully integrated Google Cloud's OKF specification into the DSOM framework to prevent AI context decay.
docs/OKF-ADOPTION-GUIDE.md & references/OKF-ADOPTION-GUIDE.md — Official guidelines for adopting OKF in DSOM.
GitBook SUMMARY.md Update. Added brand new sections for AI Agent Skills & Workflows and Genesis References.

Changed

Knowledge Bundles Schema Migration: Programmatically updated 13 .agents/brain and .agents/skills artifacts to enforce strict OKF v0.1 YAML frontmatter (title and timestamp).
Bootstrap Blueprint Portability: Converted all hardcoded absolute paths in references/HOWTO-BOOTSTRAP-SOVEREIGN-AI-PROJECT.md to relative paths for zero-lock-in portability.
GitBook Structure: Exposed the 5 Core AI Skills (including EOD/SOD Palace Sync) directly in the generated Sovereign Book.

[10.1.0-onboarding] - 2026-04-08

Added

Cross-Platform Daily Automation Rituals. Introduced Windows-native variants of Palace routines (tools/sod-palace.ps1 and tools/eod-palace.ps1) executing directly via PowerShell to bypass WSL Ansible dependencies. Symmetrically added Bash wrapper proxies (tools/sod-palace.sh and tools/eod-palace.sh) for parity on Linux/WSL2 nodes. All related documentation refactored away from explicit playbook constraints.
Smart Checkpoint Integration. Upgraded tools/checkpoint.ps1 and playbooks/dsom/checkpoint-palace.yml to support AI-generated commit messages via .agents/brain/checkpoint_summary.txt. Bridged the context gap between AI task state and human execution terminals.
Automation Hardening. Fixed non-interactive crash in reanimate.sh and reanimate.ps1 via TTY detection and robustness wrappers.
Documentation Sanitization. Purged comparative token/cost tables to maintain a neutral stance. Formally credited milla-jovovich/mempalace as the originator of the spatial memory architecture.
DSOM Onboarding Automation. Created cross-platform bootstrap scripts using Bash (dsom-onboard.sh) and PowerShell (dsom-onboard.ps1). Native Ansible fetching backend added to playbooks/dsom/onboard-dsom.yml. Handles "Zero-Debt" safe cloning into any external target repository, preventing overrides via Git branch isolation and timestamp file deduplication.
docs/HOWTO-DSOM-ONBOARDING.md documentation to guide users on "DSOM-ifying" their Git repositories fully mechanically.
README.md updated to formally include the new Onboarding Automation loop standard.

[10.0.0-palace] - 2026-04-08

Added

Sovereign Markdown Palace v1.0 — Spatial memory architecture layered on top of Git. Wing → Hall → Room → Closet hierarchy. Palace Registry as AI navigational index.
palace-sync.sh / palace-sync.ps1 v1.0 — Git history → Palace Room mapping engine. EOD batch mode (incremental) and --backfill (full history) modes.
8 Palace Rooms backfilled from full Git history (2026-01-07 to 2026-04-08): room_tooling, room_brain_artifacts, room_ledger, room_uncategorised, room_clean_architecture, room_crisp_strategy, room_dsom_protocol, room_sovereign_fabric_v9_8.
playbooks/dsom/sod-palace.yml — Ansible SOD automation. Runs git pull, audit, palace check, and reanimate. Human step (AI upload) documented explicitly.
playbooks/dsom/eod-palace.yml — Ansible EOD automation. Runs palace-sync, brain validation, selective staging, commit, and push. Closet review remains human+AI.
docs/PALACE-BUILD-STORY.md — Narrative of why and how the Palace was built.
docs/HOWTO-PALACE-ONBOARDING.md — First-time guide for users and AI agents. Includes 4-step Palace Walk protocol for AI agents.
docs/HOWTO-MIGRATE-TO-PALACE.md — 9-phase migration checklist for existing DSOM users.
docs/RESEARCH-REASONING-GAP.md — Research finding: what DSOM solves vs the live reasoning capture gap. Introduces the Decision Log Protocol.

Changed

tools/hibernation.sh → v2.1 — Step 7: Palace Spatial Reflection (auto palace-sync).
tools/hibernation.ps1 → v2.1 — Windows parity for Palace Spatial Reflection.
tools/reanimate.sh → v2.2 — Section [14]: Palace Registry in SOD manifest.
tools/reanimate.ps1 → v2.1 — Windows parity for Section [14] Palace Registry.
README.md — Updated to v10.0.0-palace. Added Palace section, Known Gap callout, updated brain structure tree, updated Key Documents table, updated roadmap.
SUMMARY.md — Added Palace Rooms, Ansible Palace playbooks, research documents.
HISTORY.md — Phase 5: Sovereign Markdown Palace & Ansible Ritual Automation added.
Documentation Fleet Sweep for Palace v1.0:
- docs/AI-MASTER-PROTOCOL.md → v6.1 + Palace v1.0. Hardcoded sod-palace.yml / eod-palace.yml routines.
- docs/CLAUDE-SETUP.md & docs/COPILOT-SETUP.md → Interwoven sod-palace.yml and spatial awareness.
- docs/CRISP2-OPERATIONAL-STRATEGY.md → Emphasized spatial context and "Zero-Debt" standard.
- docs/ITIL-ALIGNMENT.md → ITIL workflow mappings upgraded with Ansible Palace loops.
- docs/AI-RESPONSE-TEMPLATE.md → AI persona commit block updated to enforce Palace Registry checks.
- docs/HOWTO-SETUP-WSL-ALMALINUX10.md → Example code blocks upgraded to Palace playbooks.

Tagged

v10.0.0-palace — Git tag pushed to GitHub.

[9.8.0] - 2026-04-08

Added

Sovereign Persistence Fabric Finalized: Hardened 19-node Elasticsearch and Kibana fabric.
Formal Release:
[2026-04-08]: Smart Checkpoint Integration. Upgraded tools/checkpoint.ps1 and playbooks/dsom/checkpoint-palace.yml to support high-fidelity commit messages sourced from AI-generated Hibernation Notes (checkpoint_summary.txt).
[2026-04-08]: Documentation Sanitization & Crediting. Purged comparative cost data and formally credited milla-jovovich/mempalace in core protocol docs and spatial memory closets.
[2026-04-08]: Git Tag v10.1.0-onboarding pushed to GitHub. Continuous Integration loops across Windows (T1) and Linux (T2) fully harmonised.
Documentation Sanitization: Ensured Master Architect's identity and professional standards are preserved across all core docs.

[9.6.1] - 2026-04-05

Added

WSL2 Migration: Successfully transitioned Sovereign Persistence Fabric to native Linux WSL2 environment.
POSIX Compliance: Enforced strict POSIX compliance for Ansible orchestration to resolve Windows drive permission issues.
ML "Memory Inversion": Finalized ML-ready deployment architecture for 3-node standalone quorum.

[6.2.1] - 2026-04-05

Added

roles/common/ — First DSOM Ansible role. Structured with tasks/, defaults/, handlers/, meta/. Four task files:
- packages.yml — apt-based essential package installation (15 packages, idempotent).
- timezone.yml — sets Asia/Kuala_Lumpur via community.general.timezone.
- directories.yml — creates /opt/deep-state-of-mind-for-my-ai and subdirs (logs, tmp, config) owned by linuxmalaysia:1000.
- sysctl.yml — applies vm.swappiness=10 and fs.file-max=100000 to /etc/sysctl.d/99-dsom.conf.
playbooks/common.yml — Dedicated playbook to run the common role with Debian platform assertion guard.
playbooks/dsom/site.yml — Updated: roles/common now wired in (replaces inline directory task).

[6.2.0] - 2026-04-05

Added

ansible.cfg — DSOM-standard Ansible configuration. Fixed deprecated yaml callback — now uses result_format=yaml with stdout_callback=default (ansible-core 2.13+).
inventory/hosts.yml — 4-Tier topology with localhost wired as T1/T2, T3/T4 nodes template-commented for future remote targets.
inventory/group_vars/all.yml — Project-wide variables: linuxmalaysia:1000, /opt/deep-state-of-mind-for-my-ai.
playbooks/preflight.yml — Mandatory gate check playbook. Verifies connectivity, OS identity, UID, and Git. Must pass before any deployment playbook runs.
vault/.gitignore — Secrets directory protection.

Verified

ansible-playbook playbooks/preflight.yml: ok=8 failed=0 on localhost (Ubuntu 25.04, Python 3.13.3, ansible-core 2.18.1).
tools/audit-pre-flight.sh: All steps [PASS] including Ansible Baseline check.

[6.1.5] - 2026-03-29

Changed

tools/hibernation.sh — Upgraded from v1.0 (76 lines) to v2.0 (130+ lines). Replaced blind git add . with selective staging of brain artifacts + git add -u. Added: blocking walkthrough anchor check, Hibernation Notes auto-save, uncommitted file preview, privacy guardian reminder, phase-aware commit message, and improved EOD banner.
tools/hibernation.ps1 — Upgraded from v1.0 (81 lines) to v2.0 at full parity with bash. Same improvements: selective staging, blocking walkthrough check, Hibernation Notes auto-save, dirty file preview, privacy guardian reminder.

[6.1.4] - 2026-03-29

Changed

tools/reanimate.sh — Upgraded from v2.0 to v2.1. Added set -euo pipefail, colour header banner, brain artifact validation, Cognitive Twin Protocol as section [4], head-60 summaries for SOD-RITUAL and RITUAL-OF-TRANSITION, v6.1 handshake prompt in footer.
tools/reanimate.ps1 — Upgraded from v1.5 to v2.0. Achieved full 13-section parity with the bash counterpart. Added sections [4] Cognitive Twin Protocol, [12] Ansible Inventory, [13] GitOps Strategy. Fixed double-output bug from v1.5. Added brain artifact validation and Get-FileHeadSafe helper.

[6.1.3] - 2026-03-29

Changed

tools/reanimate-claude.sh — Upgraded from v1.0 (35 lines) to v2.0. Added 9 labelled sections: Sovereign Constitution, Cognitive Twin Protocol, Current Task, Mental Anchor, Implementation Plan, Git History, Project Structure, Ansible Inventory, and GitOps Strategy. Added colour output, brain artifact validation, and upload instructions.
tools/reanimate-claude.ps1 — Upgraded from v1.0 (59 lines) to v2.0 at full parity with the bash counterpart. Added structured error handling and all 9 sections.

[6.1.2] - 2026-03-29

Changed

docs/COPILOT-SETUP.md — Full rewrite to v6.1 standard. Removed duplicate Malay/English sections. Added three-pillar doctrine, SOD/EOD rituals, Stop Conditions, workspace context techniques (#file:, @workspace), .github/copilot-instructions.md and .github/prompts/dsom-reanimate.prompt.md setup guides, and cross-AI handover pointer.

[6.1.1] - 2026-03-29

Changed

docs/CLAUDE-SETUP.md — Full rewrite to v6.1 standard. Added Claude Project Instructions block (copy-paste ready), SOD/EOD ritual steps, Hibernation Notes export prompt, Stop Conditions table, cross-AI handover guidance, and Three-Pillar doctrine alignment.
docs/MULTI-AGENT-PROTOCOLS.md — Updated to v6.1. Added Google Antigravity to the Tier 1 Co-Pilot agent taxonomy.
docs/PERSONALIZATION.md — Relabelled as Gemini Edition (v6.1). Added cross-reference pointers to Claude, Copilot, and Antigravity setup docs.

[6.1.0] - 2026-03-24

Added

playbooks/dsom/site.yml and audit-preflight.yml — Established standard Ansible execution scaffolding for downstream projects.

Changed

Documentation Hardening: Full overhaul of CONTRIBUTING.md, SOD-RITUAL.md, EOD-RITUAL.md, and MIRROR-OF-KNOWLEDGE.md.
Tooling v8.6: Upgraded CheckUsage.ps1 with accurate token estimation and ASCII-safe interface.
New Resource: Created docs/tools/HOWTO-CHECKUSAGE.md.
docs/AI-COGNITIVE-TWIN-PROTOCOL.md — Removed placeholder variables and restored default baseline paths to pass the pre-flight checks.
Brain Artifacts — Updated .agents/brain/ session tracking to maintain the GitOps/AIOps continuity loop.

[6.0.0] - 2026-03-09

Added

docs/AI-COGNITIVE-TWIN-PROTOCOL.md — Generic, fillable Project Identity Card template. Defines the Cognitive Relationship, 4-Tier Environmental Map, Security Doctrine, Production Identity standard, Git Sovereignty Protocol, Execution Guardrails, Brain Sync Mandate, Advisory Mode Anchor, and Session Handover Prompt. Any new project fills in the [PLACEHOLDER] fields to adopt the DSOM Cognitive Twin model.
docs/GITOPS-AIOPS-ANSIBLE-STRATEGY.md — Three-Pillar strategic doctrine document. Defines the GitOps, AIOps, and Ansible pillars including the Integration Loop (AI Proposes → Git Records → Ansible Executes → AI Verifies), GitOps Laws, AIOps role boundaries, Ansible Laws (idempotency, no hardcoded secrets, role-based structure), and the Adoption Checklist for new projects.
docs/HOWTO-SETUP-ANSIBLE-BASELINE.md — LDP-compliant HOWTO guide for establishing the Ansible baseline in any DSOM project. Covers directory structure, ansible.cfg, inventory/hosts.yml, group variables, pre-flight playbook, Ansible Vault integration, and tools/audit-pre-flight.sh extension.

Changed

docs/AI-MASTER-PROTOCOL.md (v6.0): Added "Advisory over Execution" and "GitOps-First" sub-principles to §1; expanded §5 Coding Laws with the IaC Sovereign Law (Ansible-exclusive), Idempotency Law, and Git Sovereignty Doctrine; expanded §6 Handshake to include Cognitive Twin Protocol and Ansible Baseline verification steps.
docs/OPERATIONAL-GUIDE.md (v6.0): Updated §2 Reanimation success criteria to include AI-COGNITIVE-TWIN-PROTOCOL.md and Ansible baseline checks; added new §6 Execution Guardrails (No Silent Execution, Ansible Pre-flight Mandate, Log Review Protocol, Self-Healing Rule, GitOps Loop).
docs/REANIMATION-PROMPT-TEMPLATE.md (v6.0): Added Prompt Variant 2 — Session Handover (Cognitive Twin Trigger), the Memory Export prompt for transitioning to a new AI session, including pedagogical logic notes.
SUMMARY.md (v6.0): Added GitOps·AIOps·Ansible Strategy link under §1 Sovereign Governance; added Cognitive Twin Protocol Template under §4 AI & Agent Protocols; added new §6 Ansible & Infrastructure Automation section.
mkdocs.yml (v6.0): Added GitOps/AIOps/Ansible Strategy under Governance nav, Cognitive Twin Protocol under AI Setup nav, new Ansible nav group with HOWTO Baseline.

[5.2.0] - 2026-01-16

Added

Law 11 (ITIL 4 Alignment): Integrated Service Relationship and Value Co-creation principles into AI-MASTER-PROTOCOL.md.
Docs/ITIL-ALIGNMENT.md: New artifact defining the Service Value Chain (SVC) and SKMS.
Privacy Hardening: privacy-guardian now detects Emails, AWS Keys, and Private Keys.
Gitignore Safety: Explicitly blocked dangerous data dump formats (*.sql, *.dump).

Changed

Reanimation Handshake: Updated scripts to explicitly state the AI's role as a "Service Relationship" partner.
README.md: Added formal "ITIL 4 Service Alignment" section.

[5.1.0] - 2026-01-14

Added

Law 9 (LDP-Compliance): Integrated the Linux Documentation Project standard into AI-MASTER-PROTOCOL.md.
REANIMATION-PROMPT-TEMPLATE v1.6: Synchronised with v5.1 Master Directive, including Stop Conditions and Inward Dependency rules.
HOWTO-REANIMATE-SESSION.md: Professional user guide following LDP 'Command/Result' patterns.

Changed

Refactored docs/PERSONALIZATION.md to map directly to the CRISP-DM L1-L4 hierarchy.
Merged and consolidated multi-agent setup guides for Claude.ai and GitHub Copilot.

[5.0.0] - 2026-01-12

Added

CRISP Mandate: Established the five core operational pillars (Context, Review, Iteration, Single-purpose, Partnership).
Master Directive v5.0: Formalised the 'Sovereign Constitution' and 'Sovereign Coding Laws'.

Fixed

Improved linguistic enforcement for DBP-standard Bahasa Melayu Malaysia (Piawai) across all core artifacts.

[4.0.0] - 2026-01-09

Added

The Brain Artifacts: Initialised .agents/brain/ with task.md, walkthrough.md, and implementation_plan.md.
DSOM Persistence Protocol: Established the Handshake ritual for session reanimation.

[1.0.0] - Day 0 (2025-09-16)

Added

Initial Concept: Foundation of the Deep State of Mind (DSOM) protocol for preventing AI context decay.
Sovereign Laws: Early draft of Zero-Global Pattern and Linux-agnostic infrastructure.

Deep State of Mind (DSOM) For My AI Protocol | Harisfazillah Jamel (LinuxMalaysia) | 2026-08-21 Standard: UK English | DBP-standard Bahasa Melayu Malaysia (Piawai) | GNU General Public License v3.0