Ralph Mode: Your AI Agent That Works While You Sleep

What is Ralph Mode in MEOK AI?

Ralph Mode is MEOK's overnight autonomous agent system. You assign a mission before bed — research a competitor, draft a report, scan your logs, outline next week's content — and your Sovereign AI executes it while you sleep. By morning, a structured brief waits in your dashboard. No prompts. No babysitting. No context gaps.

The name comes from the idea of sending a trusted employee — Ralph — to handle the night shift. Unlike a human employee, Ralph never gets tired, never forgets the context from the session before, and never charges overtime. The mode is persistent, privacy-first, and entirely yours.

Who are Orion, Riri, and Hourman?

Ralph Mode is powered by three specialist agents, each with a distinct function in your overnight workflow:

• Orion. The research hunter. Orion scans sources, synthesises findings, flags threats, and builds structured knowledge briefs. Named for the hunter constellation — always searching, always returning with something.
• Riri. The builder. Riri writes code, drafts content, generates commit messages, and produces artefacts. Give Riri a spec at midnight; receive a working prototype by 7am.
• Hourman. The planner. Hourman owns your calendar, task queue, and sprint structure. It prioritises, schedules, and prepares your morning standup so you start the day with clarity rather than chaos.

The three agents coordinate through MEOK's Byzantine Council, which validates handoffs between them. Orion's research feeds Riri's builds; Hourman schedules both. The result is a coherent overnight workflow, not three isolated scripts.

What tasks can Ralph Mode complete overnight?

Ralph Mode currently supports eight task categories, each handled by the appropriate specialist agent:

• Research briefs. Orion scans web sources, academic abstracts, and news feeds to produce structured summaries on any topic you specify.
• Email drafts. Riri drafts responses to your pending emails based on thread context and your communication preferences stored in memory.
• Code commits. Riri writes, tests, and commits code to your repository against a spec you provide before bed.
• Database backups. Scheduled integrity checks and backups with a morning report on any anomalies found.
• Log analysis. Orion parses your application logs overnight and surfaces errors, patterns, and recommended fixes.
• Competitor scans. Orion monitors specified competitor domains for pricing changes, new features, and content updates.
• Content drafts. Riri produces first-draft blog posts, social content, or documentation based on your brief.
• Morning brief. Hourman synthesises all overnight work into a single prioritised briefing delivered to your dashboard at 6am.

Which pricing tier unlocks Ralph Mode?

Ralph Mode is available from the Sovereign tier at £12/month. The Sovereign tier activates all three agents — Orion, Riri, and Hourman — with full overnight task execution and morning brief delivery. The Ralph terminal interface, which lets you send natural language missions directly to the agent stack, is available to all Sovereign and above subscribers.

Free and Companion tier users get access to a limited version of Hourman for task planning, but overnight autonomous execution and Orion's research capabilities require Sovereign. If you're running a business or serious personal project, Sovereign pays for itself in the first week.

How is Ralph Mode different from ChatGPT tasks?

ChatGPT's scheduled tasks are stateless. Each run starts cold — no memory of what you asked last week, no awareness of your ongoing projects, no understanding of your communication style. They're cron jobs with a language model attached.

Ralph Mode is different in four fundamental ways:

• Persistent memory. Orion, Riri, and Hourman share access to your encrypted memory store. They know your projects, preferences, and context — and they remember them across every session.
• Care-based alignment. Every agent action passes through MEOK's care validation layer. Agents will not take actions that conflict with your stated wellbeing, even if instructed to do so.
• Personal sovereignty. Your data never trains a shared model. Your overnight work product is yours, encrypted, and portable. OpenAI trains on your interactions. MEOK does not.
• Never shared. ChatGPT task output can influence model training across millions of users. Your Ralph Mode output is private, sealed, and never leaves your encrypted vault unless you choose to export it.