Skip to content
Syn_OS Phoenix Syn_OS

GoodLife Sovereign ISO

GoodLife is the full sovereign build of Syn_OS. Same custom 7.0-synos-ai kernel, same 245-crate Rust workspace, the full ALFRED v6.0 daemon, the full AI cortex and dispatch, the complete ARCANUM mesh, and a dedicated Research Division (Jupyter, local models, ML lab) layered on top. Operational offensive tooling is simply not included — GoodLife is a complete AI-and-development platform, not an offensive-operations profile. It keeps Syn_OS’s sovereign values — no telemetry, air-gap capable, local models by default — so you get a complete, self-owned platform without becoming a target by accident. Free for personal and sovereign use; commercial or organizational deployment requires a license.

What’s in the box

Section titled “What’s in the box”
ComponentWhat you get
Kernel7.0-synos-ai with CONFIG_RUST=y, capability-gated signed Rust kernel modules — same as every other profile
ALFREDv6.0, full daemon — full AI cortex, dispatch, and consciousness fusion, with a dedicated Research Division workspace added (not a capability-restricted build)
Local LLMsOllama pre-configured with qwen2.5:7b + llama3.2:3b (~6.5 GB pre-pulled in stage 12); ISO is offline-capable
Notebook stackJupyter Lab + JupyterText + Quarto, all locally hosted
ML toolingPyTorch, TensorFlow, scikit-learn, polars, duckdb, sentence-transformers, transformers
Vector DBLanceDB local + Qdrant containerised option
RAG pipelinesynos-nlp-pipeline + synos-datalake wired into ALFRED’s hippocampus crate
Voicewhisper.cpp local transcription + Piper TTS
DesktopCinnamon DE with the GoodLife wallpaper set (goodlife-abstract, goodlife-nebula, goodlife-space)
Bevy pluginsCutscene, Mindmap, RetroFilter, Cyberspace, SkillTree, Rehoboam, Twin (FactionHQ disabled — no GRIMOIRE on GoodLife)

What’s not in the box (by design)

Section titled “What’s not in the box (by design)”

GoodLife is the full sovereign build, but it is not the offensive-operations profile. Operational offensive tooling is simply not part of the image:

  • No operational offensive tooling — no live C2, no offensive kernel modules, no operations engine. The public images carry none of it; C2 framework binaries are scrubbed at build time.
  • No GRIMOIRE gamified lab catalogue — the gamification stack is built, but the sandboxed lab catalogue and faction MMO ship on the GRIMOIRE Public profile, not here
  • No internet-facing services — sshd is disabled by default, Sanctum federation is opt-in
  • No telemetry — local-first by construction; nothing leaves your machine without explicit consent
  • No cloud LLM defaults — Ollama is the default; cloud backends (Claude / OpenAI / Gemini / DeepSeek) are off until you configure keys

The Research Division

Section titled “The Research Division”

GoodLife adds a dedicated Research Division — a Jupyter/local-model/ML workspace layered on top of the full ALFRED daemon. This is an added workspace, not a capability restriction: ALFRED runs at full capability on GoodLife (full AI cortex, dispatch, and consciousness fusion), and the Research Division is where you drive local models, notebooks, and experiments. Sovereign defaults keep everything local:

~/.config/alfred/research.toml
[backends]
default = "ollama"            # cloud backends optional, off by default
[[backends.allowed]]
name = "ollama"
[[backends.allowed]]
name = "llama-cpp"            # local fallback


[privacy]
job_hunt_mode = true          # privacy-first, no cross-context bleed
context_isolation = "tenant"  # each conversation is its own tenant

What GoodLife does not ship is operational offensive tooling. Curtain v4 holds the public images to a capability ceiling from the kernel side: GoodLife tokens cannot acquire the claims that would escalate the image to the licensed Enterprise Edition. Everything else Syn_OS can do is on the table.

Salvaged-hardware mesh

Section titled “Salvaged-hardware mesh”

GoodLife is the easiest entry point to the ARCANUM Hive: turn a stack of discarded laptops into a distributed inference mesh. The Salvage Yard quest arc lives in GRIMOIRE, but the underlying capability is a first-class GoodLife feature:

Terminal window
synos-hive-bootstrap --tier goodlife
synos-hive-add-node <mesh-ip> --role inference
synos-hive-add-node <mesh-ip> --role embedding
ollama serve --hive  # automatically shards across the mesh

A 4-node i5-class mesh outperforms a single mid-range workstation on memory-bound LLM workloads at near-zero hardware cost.

What GoodLife is for

Section titled “What GoodLife is for”
  • A complete sovereign platform — the full 245-crate stack, full ALFRED, and the full AI cortex + dispatch, all self-owned
  • Local AI research — sovereign LLM inference, no cloud bills, no telemetry
  • Federated inference experiments — the mesh is the substrate
  • RAG / embedding pipelines against your own corpus
  • Privacy-first job hunt / writing / chat — context isolation by tenant
  • The Research Division — Jupyter, local models, and the ML lab as a first-class workspace
  • Salvage operators building a sustainable home/office AI stack

What GoodLife is not for

Section titled “What GoodLife is not for”
  • Operational offensive tooling — GoodLife ships none of it; sandboxed offensive training lives on GRIMOIRE Public
  • Production multi-tenant deployment — multi-tenant ARCANUM federation and fleet management are Enterprise Edition capabilities, available under a commercial license
  • Compliance-evidence generation — the Enterprise Edition ships the FedRAMP / CMMC / SOC2 control maps and ConMon evidence collector

Getting started

Section titled “Getting started”
  1. Boot GoodLife (live or installed)
  2. First-boot wizard runs the ALFRED introduction and sets up the Research Division
  3. Open Jupyter from the desktop, or ollama run qwen2.5:7b in a terminal
  4. Optional: synos-hive-bootstrap --tier goodlife to add a second node

Next: Included Tools →

Section titled “Next: Included Tools →”
☕ buy me a coffee