Resources
Developers

Thunk

Sovereign Personal AI

A cognitive architecture that runs entirely on your hardware. Your data stays yours. Your AI answers to you.

What is Thunk?

Thunk is the flagship application built on the Thinking System framework. It's a personal AI that combines local-first execution, verifiable behaviours, and a peer-to-peer fabric that belongs to everyone.

Unlike cloud-based AI services, Thunk runs on your own hardware. Your conversations, data, and cognitive processes remain under your exclusive control. There's no silent data collection, no training on your conversations, no third-party access.

What is Thunk? Thunk is the flagship application built on the Thinking System, and it runs on your own hardware: local models, storage and inference. You reach it through ThunkUI or the API. A directive is captured as an utterance (IN_SCRIBE), its intent recognised (SUPER_SCRIBE), a workflow selected by capability, a turn contract set, and the workflow executed within a gas budget, so your conversations and cognitive processes stay with you. Every step is an envelope on the event bus, ed25519-signed, Blake3-hashed, linked to its cause and stored once in the event store, which holds your data; replay re-checks it for integrity and drift. The node joins a peer-to-peer fabric that belongs to everyone over QUIC with TLS 1.3 and ed25519 node keys, discovered via relays and DNS, with Proof-of-Performance staking. THUNK · BUILT ON THE THINKING SYSTEM Local-First Execution Local models, storage and inference run on your own hardware Your conversations Your cognitive processes You ThunkUI · API Event bus Event store Your data Utterance IN_SCRIBE Intent SUPER_SCRIBE Workflow Capability match Turn contract Grounding Execution Gas-budgeted Verifiable behaviours Every envelope is ed25519-signed, Blake3-hashed, linked to its cause and stored once. Replay re-checks the event store for integrity and drift. Peer-to-peer fabric Belongs to everyone QUIC TLS 1.3 · ed25519 node keys Discovery via relays and DNS Proof-of-Performance staking What is Thunk? Local-First Execution Thunk, built on the Thinking System, runs on your own hardware; you reach it through ThunkUI or the API. A directive is captured as an utterance (IN_SCRIBE), its intent recognised (SUPER_SCRIBE), a workflow selected by capability, a turn contract set and the workflow executed within a gas budget, so your conversations and cognitive processes stay with you. Every step is an envelope on the event bus, signed, hashed, linked to its cause and stored once in the event store, which holds your data; replay re-checks it for integrity and drift. THUNK · BUILT ON THE THINKING SYSTEM Local-First Execution Local models, storage and inference run on your own hardware You, via ThunkUI or the API Your conversations Your cognitive processes EVENT BUS EVENT STORE Your data Utterance IN_SCRIBE Intent SUPER_SCRIBE Workflow Capability match Turn contract Grounding Execution Gas-budgeted Verifiable behaviours Every envelope is signed, hashed, linked to its cause and stored once. Replay re-checks integrity and drift. What is Thunk? Peer-to-peer fabric Your node joins a peer-to-peer fabric that belongs to everyone, over QUIC with TLS 1.3 and ed25519 node keys, discovered via relays and DNS, with Proof-of-Performance staking. Peer-to-peer fabric Belongs to everyone QUIC TLS 1.3 · ed25519 node keys Discovery via relays and DNS Proof-of-Performance staking

Core Principles

Sovereignty

Autonomy over your processes and data. No external party can compel behaviour that contradicts your configured policies.

Transparency

All cognitive processes are introspectable and auditable. Every persistence decision is recorded with full provenance.

Modularity

Loosely coupled, independently verifiable components. Extend through Flows, WASM modules, and custom adapters.

Emergence

Complex behaviour from simple components. Intelligence emerges from composition, not monolithic models.

Visualising Thunk

Capabilities

Autonomous Code Synthesis

Generate, validate, and self-heal code from natural language specifications. The system iterates against a validation loop, repairing failures until it converges or exhausts its budget.

Structured Memory

Multi-tier memory with episodic, semantic, and short-term layers. All data carries bitemporal timestamps and provenance.

Declarative Workflows

Define reasoning patterns as Flows—graphs of semantic blocks transpiled to deterministic bytecode.

Policy-Governed Persistence

Only trusted Scribes can commit to long-term storage, each evaluating requests against programmable policies.

Peer-to-Peer Network

Optional participation in a peer-to-peer compute fabric (GTR-Fabric): a task marketplace with a reputation ledger and dynamic pricing. Contribute spare cycles and draw on distributed resources as the network grows.

Use Cases

Personal Assistant

Deploy on your personal hardware for a truly private assistant. Manage schedules, finances, health records—all under your exclusive control.

Enterprise

Each department deploys Scribes with specific governance policies. Cross-functional collaboration without compromising data sovereignty.

Development

Accelerate implementation with the generate-validate-heal loop. Natural language to working code with full provenance.

Research

Access decentralised compute for large-scale experiments. Publish results with cryptographic provenance.

Recommended Resources

  • Recommended: 8+ cores, 32 GB RAM, 100 GB SSD. Local models run on Apple Silicon's GPU via Ollama; more RAM lets you run larger local models.
  • Packaged app: macOS 12+ on Apple Silicon (arm64).
  • From source: the Rust workspace builds and is CI-tested on Linux (x86_64) and macOS. Linux and Windows desktop packages are not yet shipped.

Ready to Own Your AI?

Join the movement for sovereign intelligence.