Skip to main content
Live infrastructure — updated August 16, 2026

The Stack.No Smoke.

This isn't a list of tools I read about. This is what's running right now — the models, services, and infrastructure I use to build everything you see on this site.

25
AI Agents
33
Scheduled Automations
24
Cloud Models
68K+
Indexed Code Nodes
9
Active Services
100+
Skills Library

Models

24 cloud models, routed by task type

GLM-5.2

Ollama Cloud

Primary — main conversation + interactive tasks

toolsthinking

Kimi K2.7 Code

Ollama Cloud

Coding sub-agents + delegate tasks

visiontoolsthinking

DeepSeek V4 Flash

Ollama Cloud

Fallback — coding + robust tasks

toolsthinking

Gemma 4

Ollama Cloud

Vision + image analysis + goal-mode judge

visiontoolsthinkingaudio

Minimax M3

Ollama Cloud

Content drafting + creative tasks

toolsthinking

Granite 4.1

Ollama Local

Local — offline + emergency

toolsthinking

Services

All running on a single WSL2 box, 127.0.0.1 only

Hermes Gateway

:8642

AI orchestrator + Telegram bot

Python, systemd

FCC Proxy

:8082

Claude Code → cloud model routing

Python, uvicorn

Ollama

:11434

Local + cloud model inference (24 models)

Go, systemd

MemPalace

Persistent memory + knowledge graph

Python, MCP

GitNexus

Code knowledge graph (68K+ nodes indexed)

Node.js, CLI

Open WebUI

:3000

Browser-based chat interface

Docker

Open Design

:4000

Design AI stack + daemon

Node.js, systemd

Penpot

:9001

Self-hosted design platform (Figma alt)

Docker

Postiz

:4007

Social media scheduling (14 platforms)

Docker

Toolchain

What each layer does

Orchestration

  • ▸Hermes Agent v0.20.1
  • ▸delegate_task subagents
  • ▸Loop + Goal Mode
  • ▸33 cron jobs

Coding

  • ▸Claude Code (via FCC proxy)
  • ▸GitNexus code graph
  • ▸Agent-Skills (24 SDLC skills)
  • ▸Cline CLI

Memory

  • ▸MemPalace MCP
  • ▸Semantic search
  • ▸Knowledge graph
  • ▸Session diary

Models

  • ▸24 Ollama Cloud models
  • ▸Ollama Cloud (primary)
  • ▸16-model fallback chain
  • ▸Granite 4.1 (local)

Infrastructure

  • ▸WSL2 Ubuntu
  • ▸Systemd services
  • ▸16 Docker containers
  • ▸iptables firewall

How It Actually Works

A message comes in via Telegram. Hermes (the orchestrator) reads it, decides what to do, and either handles it directly or delegates to a coding subagent running through the FCC proxy. MemPalace remembers everything across sessions. GitNexus maps the codebase so changes are surgical, not guesswork. All inference runs through cloud models — 24 models, no per-token costs.

This is what's running right now. Want to see what it can do for your business?

Config backed up to github.com/joblas/Lurkr-Jo-Blade-Hermes · Last sync: August 16, 2026