Autonomous Engineering Workstation — Active Alpha

NEON SOVEREIGN

Give it a brief. Walk away.
Return to working, verified, buildable software.
From a laptop GPU to a data-centre rack — it detects your silicon and scales automatically.

✓ Check your inbox — click the link to confirmNothing in a few minutes? Check junk or spam. Outlook and Hotmail often filter new senders; Gmail usually does not.

No subscription required to operate. Perpetual licence model.

Native C++ workstation — not Electron, not a plugin. A Software House swarm runs a Task DAG,
verifies in shadow buffers, then solidifies. Local Ollama. PostgreSQL memory. Your machine.

// Why join the waitlist

Get the alpha first. Pay once, own it forever. Run it entirely on your hardware.

Waitlist members receive early install access, launch pricing before public release, and updates while the workstation is in active alpha — plus a say in what we ship next.

8 GBMinimum GPU
2 TB+Scales To
C++23Native Core
0Cloud Calls Required
AUTOAgent Count

// What it actually does

Outcomes first.
Not assistance.

Neon Sovereign is an autonomous engineering workstation: you define the outcome, it plans, executes, verifies, then hands back a buildable result with an audit trail.

Build projects from a brief

Turn a spec into a repo: scaffold, implement, compile, smoke test. You return to green builds, not a chat transcript.

Fix compiler errors

Iterative repair loops with verification gates. The ledger remembers prior incidents and proven fixes.

Modernise legacy code

Refactors with interface negotiation and Crucible verification before anything reaches disk.

Refactor large codebases

Multi-agent parallel lanes, stable semantic IDs, and rollbackable checkpoints — designed for non-trivial repos.

Generate docs that match reality

Documentation grounded in your code + build logs, not marketing boilerplate.

Run local-first, air-gapped

Local Ollama, local memory, no cloud required to operate. You control egress.

Smoke test evidence// real logs, not mockups
First successful autonomous smoke-tests

Neon Sovereign took a brief (“minimal Win32 C++20 counter app”), generated the repo (include/, src/, CMakeLists.txt), ran an integration build, verified outputs, and returned a SUCCESS verdict.

Simulation #294: Final State: SUCCESS · Seed: Completed · SwarmIdle: Yes · OutputsOK: Yes · IntegrationOK: Yes · wall_s ≈ 195s
Simulation #293: Final State: SUCCESS · Seed: Completed · SwarmIdle: Yes · OutputsOK: Yes · IntegrationOK: Yes · wall_s ≈ 4495s

Final State: SUCCESS
Seed: Completed | SwarmIdle: Yes | OutputsOK: Yes | IntegrationOK: Yes
What this proves// from the same runs
End-to-end autonomy (real tool execution)

The swarm produced buildable Win32 artifacts (CMake + C++20 source layout) and passed integration verification — without a human hand-editing files mid-run.

More demonstrations will land as the workstation UI hardens — but the build pipeline is already closing the loop.

Early proof points
2 smoke tests: SUCCESS (#293, #294)
~195s end-to-end run (#294)
0 cloud calls required
Native C++ core

// Not another copilot

Chat plugins stop.
Neon finishes.

Built from first principles for teams tired of babysitting every file edit — and architectures that cannot converge on shippable software.

Electron shell, cloud API, one assistant
Native Slint UI, local Ollama, parallel Software House
Context dies when the chat scrolls away
PostgreSQL ledger — fixes, AST graph, terminal history
Paste code, hope it compiles
Crucible gauntlet — shadow LSP, peer review, then disk
Same model writes and reviews code
Logic + Sentinel pairing enforced — critic ≠ coder
Blind to the UI it built
Silicon Retina — live framebuffer + VLM layout audit

// Hardware Detection & Auto-Scaling

Neon Sovereign reads your silicon
and scales to fill it.

At launch it queries your GPU, measures available VRAM, selects the right model tier, and sizes the agent swarm accordingly. You configure nothing. Whether on a laptop or a rack of H100s — it simply runs harder.

Laptop Class
8 GB
RTX 4060 Laptop / Mobile Workstation

Gemma-class Logic tier + compact Sentinel critic. Full pipeline, full Crucible. Audit hot-swap evicts Logic before locking Sentinel — you never pay for two full stacks at peak VRAM.

FULL CAPABILITY
Desktop / Pro
12–24 GB
RTX 3090 / 4090 / Pro GPU

Mid-tier Architect models, larger context windows, full Silicon Retina visual audit. More parallel Developer lanes active simultaneously as headroom allows.

EXPANDED SWARM
Data-Centre / Cluster
2 TB+
Multi-Host GPU Fabric / vLLM Cluster

Multi-host Ollama fabric, optional vLLM and OpenAI-compatible endpoints, rack sidecar routing. The swarm agent count grows automatically to match available inference capacity.

LIMITLESS SCALE
AUTO
No VRAM configuration. No model selection. No agent count tuning.
Neon Sovereign probes your hardware at startup, selects the appropriate model tier, applies a governor reserve, and sizes the swarm to what your silicon can sustain — from a single 8 GB card to a multi-host cluster with terabytes of VRAM.
// The Software House Pipeline

You write the brief.
Neon handles everything else.

01
Brief

State the outcome. The Architect produces a full JSON Task DAG with stable semantic IDs.

02
Plan

Kahn's algorithm computes optimal parallel execution order. You approve or steer before work begins.

03
Execute

Developer agents run in parallel, negotiating interfaces before writing a single line of code.

04
Verify

Every patch runs the Crucible gauntlet — shadow compile, peer review, adversarial audit — before touching disk.

05
Handoff

Running software, proof-of-work screenshots, forensic audit log. You return to a result, not a suggestion.


// Command Center

CHAT. BUILD. DEBUG.
One workstation.

Three intents drive the same engine — consultation when you need it, autonomous execution when you don't, forensics when something breaks.

CHAT
CONSULTATION

Talk through design and trade-offs. Swarm telemetry, staff mail, and checkpoints in one orchestrator feed — without kicking off a full build.

BUILD
AUTONOMOUS PIPELINE

Interview, plan approval, then the full Software House run: Architect DAG → parallel Developers → Crucible. Same path as headless CI.

DEBUG
FORENSICS

Reason over the open codebase and ledger — compiler incidents, prior fixes, flight-recorder history. No greenfield detours.

SPECPLANBUILDREVIEWSHIP

Pipeline lifecycle stages map to the UI and to --headless on your CI box — archived logs, staff mail, attestation chain included.


// The Agent Swarm

Role-locked agents.
Hard boundaries.
Parallel delivery.

Every agent has a fixed role enforced in C++. Wrong tools, wrong outputs — rejected at the gate. The Sentinel is always a different model family from the Architect by doctrine, so the critic never shares failure modes with the coder. The swarm count grows automatically with available VRAM.

Architect

Designs the Task DAG. Plans and arbitrates interface conflicts. Never writes implementation code.

Developer

Implements in a Reason-Act-Observe loop. Real tool execution — reads, edits, compiles, observes output.

Sentinel

Adversarial auditor. Different model family — never generates code. Red-team hypothesis on every patch.

ForgeMaster

Owns integration builds, cross-compilation, the shadow buffer. Only role permitted to call build_project.

Binary Analyst

Ghidra headless bridge. Silicon-level read-only analysis of instruction scheduling and memory layout.

Chaos Agent

Agentic fuzzing and stress testing against virtualised QEMU and Hyper-V targets.

Silicon Retina

Captures live desktop / VM framebuffers. Vision-language audit for layout collisions, contrast failures, missing UI.

Observer

Read-only blast-radius analysis. Can hold tasks and escalate before bad mutations land.

Research

Read-only reconnaissance — docs, headers, symbols — before Developers commit to interfaces.

DevOps

Environment, packaging, and deployment lanes. Coordinates with ForgeMaster on integration builds.

UX Specialist

Layout and interaction feedback paired with Silicon Retina evidence — not guesswork in chat.

Autopsy Agent

Post-mortem on failed runs — traces, logs, and ledger state to explain what broke and why.


// Project memory

It remembers
what you shipped.

LLMs are consumers of state — not the source of truth. Neon’s deterministic spine (semantic graph, task board, causal ledger) survives every session.

PostgreSQL ledger

Build history, AST vault, terminal flight recorder, FTS + pgvector recall. Bundled Postgres or your own SOVEREIGN_DATABASE_URL.

Semantic spine

Tree-sitter graph with stable semantic IDs — not fragile byte offsets. Architects emit Task DAGs that reference real nodes in your codebase.

Compiler incident memory

LNK2019, C2146, and friends get upserted. The next repair prompt starts from proven fixes, not fresh hallucinations.

Operator steering

Staff mail and blackboard replies let you redirect the swarm mid-flight without restarting the pipeline from zero.

External agents (MCP)

ledger-mcp stdio server exposes the same Postgres memory to Cursor and other tools — one ledger, many clients.

Ephemeral gossip

In-memory SQLite contract negotiation so parallel Developers agree on interfaces before anyone writes implementation code.


// The Crucible — Verification Gauntlet

Nothing reaches disk
without passing every gate.

Every patch a Developer submits runs a layered gauntlet before the filesystem sees it. Multiple inner repair rounds fire automatically before a single swarm-level retry is consumed.

01
Bracket Preflight

Fast structural balance check on the patch. Failure triggers a Speed-tier LLM structural fix before proceeding.

FAST GATE
02
Peer Gate

A second model pass, independent of the Developer, issues PEER_APPROVED or PEER_ISSUES with specific bullets. Issues feed a Logic-tier rewrite before Sentinel sees the patch.

PEER REVIEW
03
Sentinel Adversarial Audit

Red-team hypothesis plus Speed-tier audit from a different model family. Only the [AUDIT]: section drives rejection — protecting against false positives.

HARD GATE
04
Shadow Verification

LSP diagnostics and optional single translation-unit compile run against the patch in a shadow buffer. Code does not reach disk until this passes.

COMPILER GATE
05
Silicon Retina

A Vision-Language Model captures live framebuffers from the running target via QEMU, Hyper-V, or ADB. Audits for layout overlaps, contrast violations, and missing UI elements.

VLM AUDIT
06
Attestation Chain

SHA-256 attestations for every real executed command. Agents cannot forge sovereign records, golden fixtures, or integration results. Pipeline success requires a valid attestation chain.

ANTI-CHEAT

// Under the hood

Engineered in C++.
Rendered in Slint.

v0.1.0 active alpha — Windows 11 primary, WSL2/Linux build path. Agents-cpp SDK, libgit2, Tree-sitter structural edits, optional AVX2/AVX-512 dispatch.

Slint MDINative UI
OllamaManaged local
Tree-sitterStructural edits
Hyper-V / QEMUVM test lanes
Headless CI--headless pipeline
SkillsSKILL.md inject

// Zero-Leak Mandate

Fully air-gapped.
Your weights.
Your machine.
Your ledger.

Designed from the ground up for regulated and sensitive environments — healthcare, finance, legal, defence. Your source code and context never leave the box unless you explicitly configure it to.

[AI]
Local Inference by Default

Weights run on your hardware via Ollama. No cloud API key required to operate the full pipeline. Optional cluster endpoints for teams with dedicated inference infrastructure.

[DB]
Local Memory — No Cloud Sync

Project knowledge, build history, compiler incident memory — stored in a local PostgreSQL ledger with pgvector recall. No third-party knowledge base.

[ID]
Offline Identity

Collaboration uses BIP39 12-word mnemonic identity. No accounts. No registration. No cloud authentication required.

[SEC]
Encrypted Secrets

API keys and credentials encrypted via Windows DPAPI. Never exposed to any model context window under any circumstance.

// ZERO-LEAK STATUS
cloud_api_requiredFALSE
weights_locationLOCAL
source_code_egressNONE
project_memoryLOCAL POSTGRES
secrets_exposureDPAPI ENCRYPTED
identity_modelBIP39 MNEMONIC
vendor_accountNOT REQUIRED
air_gap_deploySUPPORTED
regulated_envSUITABLE

// Universal Target Scope

The swarm adapts
to your target.

Toolchain, verification strategy, and agent specialisation change automatically based on what you are building. Tier-1 depth on native C/C++/Rust today — everything else in active expansion.

CC++23Rust ZigTypeScriptReact VueWASMSwift (iOS) Kotlin / NDKARM Cortex-MRISC-V AVRESP32Unreal C++ Unity C#GLSL / HLSLCOBOL FortranAdax86 ASM ARM ASMMIPS ASMQEMU — Solaris QEMU — IRIXVAX / VMSPython via WSL Node via WSLGo via WSL

TIER-1 (highlighted) — C / C++ / Rust: stable, production, battle-tested. All others in active expansion.

// System Requirements & Alpha Status

Active Alpha.
Honest limitations.

Neon Sovereign is shipping to early access registrants first. It is a genuine engineering frontier — not a polished consumer product. Expect rapid iteration and breaking changes.

OS
Windows 11 (23H2+)
WSL2 Ubuntu 22.04+
Version
v0.1.0 Alpha
GUI beta — rapid iteration
Build Tools
Visual Studio 2022
CMake 3.28+, Rust, Git
GPU (minimum)
8 GB discrete VRAM
Auto-scales to 2 TB+
Inference
Ollama on-box
Optional cluster / vLLM
Tier-1 Targets
C / C++ / Rust
Web & Mobile stabilising
Virtualisation
Hyper-V requires
Windows Pro/Enterprise
// Early Access

Define the outcome.
Neon Sovereign handles the execution.

Native C++ workstation — parallel agents, shadow-buffer verification, persistent ledger. Not a chat sidebar. Join the waitlist for alpha access.

✓ Check your inbox — click the link to confirmNothing in a few minutes? Check junk or spam. Outlook and Hotmail often filter new senders; Gmail usually does not.

NeonSovereign.store  ·  Perpetual licence. No subscription to operate.