Important: this series explores the broader Nano thesis. It is not the language specification, an API reference, or evidence that a capability is implemented. For current behavior, start with the architecture, language reference, and status.
These short essays record the rationale behind the project, possible future directions, and trade-offs worth testing. Some arguments intentionally extend beyond the alpha Python reference runtime. When a paper differs from the code or reference documentation, the code and tests define the current contract.
| # | Paper | Question it explores |
|---|---|---|
| 01 | Why Nano | Why might deterministic compiled rules complement model-driven systems? |
| 02 | The Agentic Compiler | What would it mean to compile decisions rather than invoke a model every time? |
| 03 | Determinism | Why does deterministic replay matter for a reference runtime? |
| 04 | Provenance: Protocol C | How could an optional host integration add signed decision receipts? |
| # | Paper | Question it explores |
|---|---|---|
| 05 | Why Not Go? | Where might a narrow decision DSL fit beside general-purpose languages? |
| 06 | Why Not TypeScript? | What is different about modeling constrained decisions? |
| 07 | Nano vs. Pine Script | How do the strategy examples compare with indicator-oriented scripting? |
| 08 | LLM Integration | What could model integration look like? (Not implemented in Nano v0.1.0.) |
| 09 | Autonomous Systems | Which broader system shapes motivate the project? |
| 10 | The Nano Family | What is the proposed Nano / Nano+ / Nano++ framing? |
| # | Paper | Question it explores |
|---|---|---|
| 11 | Performance | What should future benchmarks measure? |
| 12 | Security | Which constraints can the strategy IR enforce, and where are host boundaries? |
| 13 | Design Principles | Which architectural constraints should guide future work? |
| 14 | Future Work | Which capabilities are still unimplemented or experimental? |
| 15 | Closed-Loop Engineering | What might governed self-improvement require? |
| 16 | Quantum Computing | When might specialized optimization backends be worth researching? |
| 17 | Heterogeneous Compute | What could portable execution across backends mean? |
The shipped repository includes a small strategy DSL, serializable strategy IR, deterministic reference interpreter, host decision-gate bridge, test corpus, and several experimental primitives. It does not include an LLM runtime, automatic escalation, general loop executor, live exchange integration, or real quantum dispatch. Use status.md as the concise capability map.