Private AI Operations Environment
by Mahfuzur
Private AI operations environment for secure automation, orchestration, research, and enterprise workflows.
Access is required before using KernelX engines.
Authenticated Command Surface
Logged in as unknown
AI Path Academy
Manage publication Tasks without exposing Engines, credentials, or execution internals.
Open Publication to load operational status.
Loading engine lifecycle status...
Checking engine status...
Task Status
Checking task status...
Server Status
Checking server status...
Command Queue
Loading Action Center status...
Message of the Day
KernelX should generate reusable task logic, prompts, rules, blueprints, and learning notes.
No approval queue yet. This will appear after task persistence and live monitoring are added.
ACTIVE ENGINE
Manage Email Monitor Engine tasks. No Gmail, n8n, provider, or external action is executed yet.
Generate a task config from a high-level instruction, then save it to KernelX.
Sources are loaded from Settings → Email Connectors.
Channels are loaded from Settings → Notification Channels.
These are saved as task metadata only. No provider login, mailbox read, n8n call, draft creation, email send, or external execution is started.
No preview generated yet.
Click a task row to open the Task Console. Use View Saved Config to see the raw saved task configuration.
This shows the config already saved in Postgres.
Review, approve, reject, and safely execute proposed actions from any KernelX engine.
Governed Engine Factory
Preview a future KernelX engine blueprint, check required adapters, and keep creation/deployment disabled until approval.
Engine Blueprint Preview
Factory Boundary
Blueprints vs Factory
Blueprints are for planning and reusable designs. Engine Factory is the governed creation pipeline for real KernelX engines.
Governed Engine Registry
Inventory and lifecycle control for active, generated, archived, and roadmap engines.
Active / Reference
Implemented or reference-pattern engines.
Disabled
Engines disabled from runtime use.
Archived
Engines retained for history and audit.
Roadmap
Future engines requiring Factory and Action Center approval.
Planned Engine
Future engine for creating task config, connector blueprint, prompt rules, test examples, and execution preparation.
Command Center
Telegram-like action inbox for notifications and approvals from KernelX engines.
No messages loaded yet.
KernelX Settings
Engine preferences, provider readiness, connector status, and runtime safety controls.
LLM Foundation
Manage saved LLM provider and router metadata. Secret values are stored through KernelX Vault only. No external model call is made yet.
Add LLM provider
KernelX Vault Secret
Store the real API key or router token in KernelX Vault. The raw value is never returned after save.
Email Foundation
Manage saved email connector metadata. Multiple accounts are supported. Raw passwords, OAuth tokens, and API secrets are not stored here.
Add connector
KernelX Vault Secret
Store the real app password or token payload in KernelX Vault. The raw secret is never returned after save.
Notification Foundation
Manage where KernelX can notify you. Command Center is active now. Telegram, Discord, and webhooks are settings-only until explicitly connected later.
Add notification channel
Execution Bridge
n8n is only a trigger, connector, and execution bridge. KernelX, OpenClaw, and Hermes stay in control. No workflow is created, updated, called, or executed yet.
Automation Factory
Preview how Hermes would design an automation workflow while KernelX/OpenClaw keep control and n8n remains only the execution bridge. Nothing is created or executed.
Recent saved blueprint previews. These are preview records only. No LLM, provider, n8n, workflow, or external execution was triggered.
Governed Engine Factory
Preview a future KernelX engine blueprint, check required adapters, and keep deployment disabled until approval.
Runtime Boundary