emergE ARCHITECTURE

Purpose

Human-readable knowledge base with Markdown files as the source of truth for strategy docs, meeting notes, research, and operational context. Each venture has its own vault with standardized folder structure.

Architecture

• Local-first: all vaults stored on-disk as plain Markdown
• Version control per vault via Git
• Indexed by Knowledge Graph Engine for hybrid vector + full-text search
• Cross-vault connections discovered by Knowledge Graph Connection Finder

Integration Points

• Knowledge Graph Engine syncs vault pages into the graph
• Re-embeds changed pages automatically
• Structural analysis across vaults via scheduled jobs
• Capture endpoint auto-routes thoughts/ideas to correct vault

Purpose

Skills are reusable, composable instruction sets that teach the agent how to perform specialized tasks. Each skill is a directory containing instructions plus optional scripts, reference files, and config. They load on-demand when the agent detects a matching task.

Skill Categories

• Build: Multi-stage code pipeline, build orchestration sub-skills
• Research: Browse & learn, deep inquiry, content recycling
• Operations: Daily briefing, health checks, full & git backups
• Communication: Social media posting, engagement, URL shortening
• Knowledge: Knowledge graph management, enhanced memory
• Infrastructure: Platform ops, session logs, node connectivity

Resolution

When a task arrives, the agent scans all skill descriptions. If exactly one matches, it reads the instructions and follows them. Skills are never stacked — only one per task.

Purpose

Bridges the gap between session-based memory (volatile) and permanent storage (knowledge base vaults). Maintains a living model of the current operational state — what's being worked on, who's involved, and how entities connect. This allows the agent to resume context after compaction or restart.

Components

• Entity state: Tracks all known people, companies, projects, and their current status. Updated after every significant interaction.
• Active threads: WIP tracker for unfinished tasks, pending decisions, and open conversations across channels.
• Cross-reference index: Bidirectional entity relations. Enables "who works on what and where is it documented" queries.
• Session snapshots: Full state capture before context compaction so no context is lost.

Purpose

Three-tier memory system providing different persistence guarantees: daily logs for raw events (append-only), curated long-term memory (manually maintained), and disaster recovery for full system restore capability.

Memory Hierarchy

• Working memory: Current session context window (compacted via local models)
• Daily logs: Raw chronological events, append-only during pre-compaction flush
• Long-term: Curated, deduplicated, human-readable summary of permanent knowledge
• Disaster recovery: Rolling 3-backup rotation for full system restore

Purpose

The default model for all agent operations. Handles task execution, conversation, tool orchestration, and general reasoning. Selected for strong multi-step reasoning, fast response times, and cost efficiency.

When It's Used

• All standard agent interactions across channels
• Sub-agent spawning for parallel task execution
• Code pipeline stages (architecture, testing)
• Daily operations, briefings, and routine decisions

Purpose

Premium model for tasks requiring deeper analytical capability. Used as fallback when the primary model encounters complex reasoning challenges, for code review, and as an independent perspective in debates.

When It's Used

• Complex multi-step reasoning tasks
• Code review in the multi-stage pipeline
• Whitepaper debate partner (independent perspective)
• Fallback routing when primary model fails

Compaction Role

A lightweight variant handles context compaction — summarizing long conversations to fit within token limits. This is a cost-saving measure for summarization tasks.

Purpose

Deep research and analysis model. Used for tasks requiring extensive web search synthesis, long-document analysis, and cross-domain reasoning.

When It's Used

• Industry research and competitive analysis
• Long-document summarization and synthesis
• Cross-referencing multiple sources for due diligence
• Research tasks that benefit from its large context window

Purpose

Visual asset generation for all ventures. Produces illustrations, social media graphics, UI mockups, and presentation assets on demand.

Integration

• Called via the agent's image generation tool
• Supports text-to-image, image editing, and style transfer
• Output automatically saved to managed media directory
• Used for social media post images, diagrams, pitch deck assets

Purpose

Primary communication hub for all operations. Each venture has a dedicated channel for focused context. The platform also hosts a bridge service that syncs issue status between messaging and the project management backend.

Channel Architecture

• Infrastructure — System ops, alerts, gateway status
• Brand & Marketing — Personal brand, social media management
• Trading — Trading platform development
• Agency — Automation agency, client work
• Marketplace — Marketplace operations

Co-Worker Access

Trusted co-workers interact via scoped permissions — read access with sandboxed workspace, or full technical access with project management CRUD depending on role.

Purpose

Direct line to leadership for time-sensitive alerts, daily briefings, and quick interactions. Used for heartbeat notifications when the system detects urgent items requiring human attention.

Message Types

• Heartbeat alerts: Urgent emails, events <2h away, prolonged silence detection
• Daily briefings: Morning summary of overnight activity, pending decisions, schedule
• Quick commands: Short instructions from mobile
• Approval flow: External action confirmations (email drafts, social posts)

Rules

• Quiet hours: no outbound unless urgent during off-hours
• No markdown tables — use bold text or CAPS for structure
• Batch information — no drip-feeding

Purpose

Browser-based control panel for the Agent Orchestrator. Provides session management, configuration editing, tool testing, and real-time agent monitoring. Used primarily for debugging and administration.

Features

• Session listing and history inspection
• Agent configuration and model overrides
• Tool testing and approval management
• Gateway status and health monitoring
• Scheduled job management

Purpose

Cron-driven automation for recurring tasks. Ensures the system self-maintains without human intervention — backups, health checks, knowledge updates, and cleanup all happen on schedule.

Scheduled Jobs

• System backup: Weekly full backup to external drive
• Git backup: Daily version-controlled backup of critical state
• Health check: Every 15 minutes — verifies all services are running
• Knowledge distiller: Weekly structural analysis across knowledge base
• Knowledge graph sync: Every 6 hours — re-imports and re-embeds stale content
• Context snapshot: As needed before compaction events

Purpose

Continuous monitoring of all system services with automatic recovery. Each service runs as a managed daemon — if it crashes, it restarts automatically within minutes.

How It Works

• All critical services run as managed launch daemons
• Health check script runs every 15 minutes
• Verifies each service endpoint is responding
• Auto-restarts any service that's down
• Alerts sent to messaging channels on repeated failures

Purpose

Background agent running on a local model to handle project management issue updates — status changes, priority sorting, and automated issue triage. Offloads routine management from the primary agent.

Responsibilities

• Issue status synchronization
• Priority-based sorting and labeling
• Routine status updates to messaging channels

Purpose

Browser-based terminal for remote system administration. Allows authorized users to access the workstation's command line from any device without SSH setup.

Capabilities

• Full shell access via web browser
• Real-time command output streaming
• Useful for emergency debugging from mobile
• No VPN/SSH required — accessible from any network

Purpose

Central orchestrator for all agent operations. Routes incoming messages from channels to the appropriate agent session, manages context windows, handles tool calls, and coordinates sub-agent spawning.

Core Functions

• Session management: Each channel+conversation gets a persistent session with its own context window
• Context compaction: When context approaches token limits, local models summarize and compress older messages
• Tool routing: Maps agent tool calls to actual system operations (file read/write, shell exec, web fetch, etc.)
• Channel multiplexing: One agent can serve multiple channels simultaneously
• Sub-agent spawning: Isolate complex tasks in child sessions (debates, coding, research)
• Heartbeat system: Periodic check-ins for urgent items (email, calendar, schedule conflicts)

Startup Flow

Gateway starts → loads skills → initializes channels → restores sessions → begins heartbeat loop. All state persists across restarts via on-disk session storage.

Purpose

Structured project management across all ventures. Every task for a tracked project goes through this system — create issue, get identifier, execute, mark done.

Entity Tracking

• Enterprise automation venture
• Trading automation platform
• Marketplace venture
• Personal brand management
• AI shopping assistant concept
• Cross-venture tasks

Integration

• Messaging Bridge syncs issue status to venture channels
• Multi-Stage Code Pipeline tracks stages via issue labels
• Agent creates issues automatically when tasks are identified
• Relational database backend for persistent storage

Purpose

Searchable knowledge graph that indexes all knowledge base vaults and enables hybrid search across the entire organizational knowledge base. Powers the agent's ability to retrieve relevant context from hundreds of documents instantly.

Technical Architecture

• Storage: Embedded database — zero-config, file-based
• Search: Hybrid vector similarity + full-text search
• Embeddings: Local embedding model (free, on-device)
• CLI: Command-line interface for import, search, and maintenance

Key Operations

• Import vault pages into knowledge graph
• Re-embed changed pages only (incremental)
• Hybrid search across all vaults
• Verify index health and coverage

Purpose

Legacy agent platform with extensive built-in skills and browser automation. Currently in reserve — not the daily driver but available for tasks that benefit from its extensive skill library or Python-native tooling.

Capabilities

• Pre-built skills for various automation tasks
• Python virtual environment with full library access
• Browser automation engine
• Can be activated for specific workloads

Status

Active reserve. Superseded by the Agent Orchestrator as the primary framework, but maintained for specialized tasks and as a fallback.

Purpose

Local LLM inference engine providing zero-cost AI capabilities for high-volume, latency-tolerant tasks. Runs on Apple Silicon GPU via Metal acceleration.

Models

• Primary Compaction Model: Context compaction, safeguard checks, background tasks. Primary workhorse.
• Backup Compaction Model: Backup compaction, cognitive analysis
• Project Management Model: Dedicated model for issue management agent
• Embedding Model: Text embeddings for knowledge graph

Why Local

• Zero incremental cost — critical for high-frequency operations like compaction
• No network latency — compaction runs in seconds
• Privacy — sensitive context never leaves the machine
• Resilience — works offline, no API dependency

Purpose

In-memory data store used for caching, message queuing, and ephemeral session state. Provides sub-millisecond access for frequently-needed data.

Usage

• Cache: LLM response caching, tool result memoization
• Queue: Message queue between services
• Session state: Ephemeral data that doesn't need persistence (conversation state, temporary variables)

Purpose

Persistent relational database for structured data. Powers the Project Management System's issue tracking and the Knowledge Graph Engine's embedded storage.

Usage

• Project Management: Issues, status boards, entity tracking
• Knowledge Graph: Embedded instance for vector + full-text search
• Configuration: System-level settings and metadata

Purpose

Local model testing environment for experimenting with new models before deploying them to the inference engine. Provides a GUI for loading, configuring, and testing models.

Usage

• Test new model versions before promotion
• Benchmark model performance on Apple Silicon
• Compare model outputs side-by-side
• Not part of production pipeline — experimentation only