Skip to content

AI Slash Commands Guide (DSOM)

Entry Point 7: This document is part of the Procedural Skill Entry Point. See START-HERE.md for the master onboarding roadmap.

This document serves as the authoritative reference for invoking Antigravity slash commands to orchestrate the Deep State of Mind (DSOM) Cognitive Twin. These directives are specifically mapped to the workflows of managing the Sovereign AI Agent Workspace, cognitive architecture, and Git-native PMO operations.

Core Operational Directives

/goal (The Unattended Execution Protocol)

Purpose: Mandates the AI to execute exhaustive, multi-step operations autonomously without pausing for manual intervention at trivial hurdles. Sovereign Workspace Use Case: - Knowledge Ingestion: "Execute the dsom-knowledge-ingester skill on the provided research PDF and don't stop until the synthesized summary is fully integrated and committed into the Sovereign Markdown Palace." - Compliance Auditing: "Audit the entire .agents/brain directory to ensure all markdown files contain valid OKF v0.1 frontmatter, and automatically inject any missing signatures." Impact: Ideal for heavy-lifting cognitive tasks where persistence is required over immediate dialogue.

/schedule (The Temporal Orchestration Protocol)

Purpose: Establishes recurring task pipelines or delayed one-shot triggers directly in the background. Sovereign Workspace Use Case: - Palace Maintenance: "Run the palace-auditor skill every 4 hours in the background to ensure no dead links exist in the Sovereign Markdown Palace and alert me of any orphans." - Swarm Synchronization: "Wait 5 minutes for the subagent swarm to finish drafting the governance documents, then compile the final Markdown files into a PDF." Impact: Eliminates the toil of manual polling and allows the architect to context-switch while the AI handles the temporal logic.

/grill-me (The Interrogation Protocol)

Purpose: Forces the AI into an interactive interview mode to resolve ambiguity and formulate an airtight blueprint before execution. Sovereign Workspace Use Case: - Architecture Design: "We need to formulate a new Cognitive Protocol for handling long-term context decay. Grill me on the structure, memory tiering requirements, and how it integrates with the existing Palace Registry." Impact: Prevents the AI from making dangerous assumptions about critical knowledge architecture or core rules.

/learn (The Persistent Cognitive Integration Protocol)

Purpose: Compresses a resolved complex edge case into a universal structural rule and permanently commits it to the AI's global memory (.agents/AGENTS.md). Sovereign Workspace Use Case: - Constraint Assimilation: "We just spent 2 hours solving the UTF-8 encoding corruption caused by PowerShell's default ANSI output. Use /learn to permanently add this to your core rules so you always explicitly specify -Encoding utf8 in future scripts." Impact: Guarantees that hard-fought engineering and architectural lessons are permanently assimilated into the AI's operational baseline.

Ancillary Engineering Directives

/browser (The Web Extraction Protocol)

Purpose: Grants the AI capability to autonomously navigate the web, scrape documentation, and index external intelligence. Sovereign Workspace Use Case: - Framework Triaging: "Search the web for the latest advancements in LLM context-window management and cross-reference them against our DSOM Progressive Disclosure architecture." - Documentation Retrieval: "Pull the official MkDocs Material theme documentation and summarize how we can implement a custom CSS dark mode for the Sovereign Engine." Impact: Accelerates external research directly within the terminal context, pulling raw data for synthesis.

/teamwork-preview (The Cognitive Swarm Protocol)

Purpose: Deploys a swarm of autonomous subagents to tackle large-scale, concurrent engineering or documentation tasks. Sovereign Workspace Use Case: - Mass Auditing & Generation: "Deploy a team of subagents. Have one agent draft the new architecture blueprint in docs/governance, while another agent simultaneously builds a Python script to validate the OKF YAML frontmatter across the repository." Impact: Massively scales concurrent read/audit/write operations across a distributed workspace using Git Worktree Isolation.

Deep State of Mind (DSOM) For My AI Protocol | Harisfazillah Jamel (LinuxMalaysia) | 2026-07-17 Standard: UK English | DBP-standard Bahasa Melayu Malaysia (Piawai) | GNU General Public License v3.0