NAP — Narrative Addressing Protocol

NAP is a protocol that makes narrative resources addressable, resolvable, and interoperable across tools, storage systems, formats, and AI workflows.

Characters, locations, scenes, props, and entire fictional universes — NAP gives each one a stable URI, a human-and-machine-readable manifest, a content-addressed history, and a resolver that connects them all.

In the same way that IPFS content-addressed files and OCI container-addressed images, NAP is narrative-addressed — a universal namespace for the building blocks of stories.

Why NAP?

Today, narrative assets live in silos:

• Worldbuilding docs in Notion or Google Docs
• Character sheets in spreadsheets
• Concept art in Dropbox or S3
• Scene breakdowns in Final Draft or Fade In
• AI prompts scattered across chat logs
• 3D assets on Sketchfab or Polycam

None of these tools talk to each other. NAP unifies them under a single addressing and resolution layer.

nap://starwars/character/lukeskywalker
nap://starwars/location/tatooine
nap://starwars/scene/cantina
nap://toystory/prop/andy-hat

Core Primitives

NAP is built on four primitives:

1. URI — Identity

A nap:// URI identifies any narrative resource. Version, branch, and tag are orthogonal selectors passed alongside the URI — never encoded in the path (mirrors Git, OCI, and package managers).

nap://starwars/character/lukeskywalker#references.appears_in
────┬── ───┬──── ────┬──── ──────┬────── ─────────────┬───────────
 scheme universe  entity_type entity_id          fragment (query)

2. Manifest — Current State

A YAML manifest is the durable representation of a narrative resource. It is simultaneously:

• Human-editable — readable by worldbuilders
• Machine-editable — structured, schema-validated
• Agent-readable — subtree-queryable for AI workflows
• Portable — no runtime dependency, just a file
• Signable — hash the content, sign the hash (Ed25519 in v0+)
• Versionable — the manifest is what gets committed

id: "nap://starwars/character/lukeskywalker"
name: "Luke Skywalker"
entity_type: character
version: 17
properties:
  homeworld: "nap://starwars/location/tatooine"
  species: human
representations:
  reference_image:
    hash: "sha256:e3b0c44..."
    format: png
provenance:
  model: "midjourney-v6"
  prompt_hash: "sha256:abc123..."
head: "a72c9f3b..."

3. Commit — History

Commits are content-addressed (SHA-256) snapshots with patch metadata. The manifest stores only head — a pointer to the latest commit. Full history lives in the VCS, keeping manifests bounded.

4. Resolver — URI → Manifest

The resolver turns a nap:// URI into a manifest (or a subtree of one). With optional selectors for branch, tag, or commit hash, it supports versioned resolution and fragment-based queries for efficient data access.

Entity Types

Groups

Groups are composite objects that can reference both entities and media.

Use cases: compound props, outfits, events, and more.

Examples: woody + buzzlightyear = group("woodyandbuzzflying"). andy + truck = group("andysmovingtruck").

Repository Layout

Each universe is a Git repository on disk:

starwars/                    ← universe root (Git repo)
├── .nap/
│   └── config.yaml          ← repository configuration
├── universe.yaml            ← world manifest
├── characters/
│   ├── lukeskywalker.yaml
│   └── darthvader.yaml
├── locations/
│   └── tatooine.yaml
├── scenes/
│   └── cantina.yaml
└── props/

Quick Start

Install

CLI & Server (Rust — compile from source)

git clone https://github.com/cinematiccanvas/nap.git
cd nap
cargo build --release

# Binaries land in target/release/
#   nap          — CLI tool
#   nap-server   — HTTP resolver server

Python SDK (prebuilt wheel, no Rust needed)

pip install narrativeengine

from narrativeengine import create_block, generate_candidate, render_lore_summary

block = create_block("char-1", "A brave adventurer")
candidate = generate_candidate(block)

TypeScript SDK (prebuilt binary, no Rust needed)

npm install narrativeengine

import { createBlock } from "narrativeengine";

const block = createBlock("char-1", "A brave adventurer");

Create a Universe

# Initialize a new universe
nap init starwars

# See what you created
ls starwars/
# → .nap/  universe.yaml  characters/  locations/  scenes/  props/

Create & Inspect Entities

# Create a character
nap create character lukeskywalker -u starwars -n "Luke Skywalker"

# Create a location
nap create location tatooine -u starwars -n "Tatooine"

# Set properties
nap set nap://starwars/character/lukeskywalker species human
nap set nap://starwars/character/lukeskywalker homeworld "nap://starwars/location/tatooine"

# Resolve a manifest
nap resolve nap://starwars/character/lukeskywalker

# Query a specific field
nap resolve nap://starwars/character/lukeskywalker#properties.species
# → human

# Query a subtree
nap query nap://starwars/character/lukeskywalker properties

Version Control

# View commit history
nap history nap://starwars/character/lukeskywalker

# Create branches
nap branch starwars canon

# Create tags
nap tag starwars episode-4

Output Formats

nap resolve nap://starwars/character/lukeskywalker -f json
nap resolve nap://starwars/character/lukeskywalker -f yaml

HTTP Server

The NAP resolver server provides a REST API for resolution and commits.

# Start the server (defaults to port 3100, base path = current directory)
nap-server

# Custom port and base path
NAP_PORT=8080 NAP_BASE_PATH=/path/to/universes nap-server

Endpoints

Query parameters for resolution: branch, commit, tag, path (subtree query).

AI Workflows

NAP is designed for AI-native workflows from day one:

Subtree queries let AI agents fetch exactly the data they need — 500 tokens instead of 40,000:

nap resolve nap://starwars/character/lukeskywalker#references.appears_in
nap resolve nap://starwars/scene/cantina#properties.mood

Provenance tracking records AI generation metadata in the manifest itself:

provenance:
  model: "midjourney-v6"
  prompt_hash: "sha256:abc123..."
  seed: "42"
  derived_from: "nap://starwars/character/lukeskywalker/v1"

Content-addressed representations link manifests to assets by hash:

representations:
  reference_image:
    hash: "sha256:e3b0c44..."
    format: png
  voice_model:
    hash: "sha256:def567..."
    format: onnx

Project Structure

nap/
├── Cargo.toml                      ← workspace root (7 crates)
├── crates/
│   ├── nap-core/                   ← core library (URI, manifest, resolver, VCS)
│   │   └── src/
│   │       ├── lib.rs              ← crate root, re-exports
│   │       ├── uri.rs             ← NapUri parser/builder
│   │       ├── manifest.rs         ← Manifest, Representation, Provenance
│   │       ├── commit.rs           ← Commit, Change, ChangeOp
│   │       ├── resolver.rs         ← Resolver (URI → Manifest)
│   │       ├── query.rs            ← Subtree query engine
│   │       ├── repository.rs       ← Universe repository CRUD
│   │       ├── types.rs            ← EntityType enum
│   │       ├── content.rs          ← SHA-256 content hashing
│   │       ├── error.rs            ← NapError types
│   │       ├── vcs.rs              ← VcsBackend trait
│   │       └── vcs_git.rs          ← Git backend implementation
│   ├── nap-cli/                    ← CLI binary (nap)
│   ├── nap-server/                 ← HTTP server binary (nap-server)
│   ├── narrativeengine-core/       ← narrative engine (AI story generation)
│   ├── narrativeengine-py/         ← Python bindings (PyO3)
│   ├── narrativeengine-ts/         ← TypeScript/NAPI bindings
│   └── narrativeengine-codegen/    ← schema/code generation tooling
├── python/                         ← Python SDK package
│   └── pyproject.toml
└── typescript/                     ← TypeScript SDK package
    ├── package.json
    ├── index.cjs
    └── index.d.ts

Build & Test

Prerequisites

• Rust 2024 edition (1.85+) — only needed to build from source
• Git (for the VCS backend)

Pre-commit Hooks

This repo ships a pre-commit hook that runs fast checks (cargo fmt, ruff, eslint, vitest) before each commit. Activate it once per clone:

git config core.hooksPath .githooks

The hook only runs checks relevant to the files you've staged — no Rust checks on pure Python changes, etc.

Build

# Build everything (debug)
cargo build --workspace

# Build everything (release)
cargo build --release --workspace

# Build individual crates
cargo build -p nap-core
cargo build -p nap-cli
cargo build -p nap-server
cargo build -p narrativeengine-core
cargo build -p narrativeengine-codegen

Test

# Run all tests (excluding Python bindings which need Python headers)
cargo test --workspace --exclude narrativeengine-py --exclude narrativeengine-ts

# Run tests for a specific crate
cargo test -p nap-core

# Run doc tests
cargo test --doc

Build SDK artifacts

# Python wheel (requires maturin)
pip install maturin
cd python
maturin build --release

# TypeScript prebuild (requires napi-rs toolchain)
cd typescript
npm install
npm run build:native

Run

# CLI
cargo run -p nap-cli -- --help

# Server
cargo run -p nap-server

Design Principles

Manifest is current state. History is external.

• Manifests store only head — a pointer to the latest commit.
• Full history lives in the VCS, preventing unbounded manifest growth.

Version/branch/tag are NEVER in the URI.

• They are orthogonal selectors passed alongside the URI (mirrors Git, OCI, package managers).

Content-address everything.

• Every representation is identified by its SHA-256 hash.
• Manifests are content-hashable for signing and verification.

Subtree queries are first-class.

• AI systems, CLI tools, and HTTP clients all use the same query engine.
• Fragment queries enable efficient data access without fetching entire manifests.

Status

NAP is in v0 (prototype) — the core data model and resolution engine are functional. Signing and verification are stubbed for future iterations.

License

MIT © Cinematic Canvas