Files

giles 1a74d811f7 Incorporate art-dag-mono repo into artdag/ subfolder

Merges full history from art-dag/mono.git into the monorepo
under the artdag/ directory. Contains: core (DAG engine),
l1 (Celery rendering server), l2 (ActivityPub registry),
common (shared templates/middleware), client (CLI), test (e2e).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

git-subtree-dir: artdag
git-subtree-mainline: 1a179de547
git-subtree-split: 4c2e716558

2026-02-27 09:07:23 +00:00

3.1 KiB

Raw Blame History

artdag

Content-addressed DAG execution engine with ActivityPub ownership.

Features

Content-addressed nodes: node_id = SHA3-256(type + config + inputs) for automatic deduplication
Quantum-resistant hashing: SHA-3 throughout for future-proof integrity
ActivityPub ownership: Cryptographically signed ownership claims
Federated identity: @user@artdag.rose-ash.com style identities
Pluggable executors: Register custom node types
Built-in video primitives: SOURCE, SEGMENT, RESIZE, TRANSFORM, SEQUENCE, MUX, BLEND

Installation

pip install -e .

Optional: External Effect Tools

Some effects can use external tools for better performance:

Pixelsort (glitch art pixel sorting):

# Rust CLI (recommended - fast)
cargo install --git https://github.com/Void-ux/pixelsort.git pixelsort

# Or Python CLI
pip install git+https://github.com/Blotz/pixelsort-cli

Datamosh (video glitch/corruption):

# FFglitch (recommended)
./scripts/install-ffglitch.sh

# Or Python CLI
pip install git+https://github.com/tiberiuiancu/datamoshing

Check available tools:

python -m artdag.sexp.external_tools

Quick Start

from artdag import Engine, DAGBuilder, Registry
from artdag.activitypub import OwnershipManager

# Create ownership manager
manager = OwnershipManager("./my_registry")

# Create your identity
actor = manager.create_actor("alice", "Alice")
print(f"Created: {actor.handle}")  # @alice@artdag.rose-ash.com

# Register an asset with ownership
asset, activity = manager.register_asset(
    actor=actor,
    name="my_image",
    path="/path/to/image.jpg",
    tags=["photo", "art"],
)
print(f"Owned: {asset.name} (hash: {asset.content_hash})")

# Build and execute a DAG
engine = Engine("./cache")
builder = DAGBuilder()

source = builder.source(str(asset.path))
resized = builder.resize(source, width=1920, height=1080)
builder.set_output(resized)

result = engine.execute(builder.build())
print(f"Output: {result.output_path}")

Architecture

artdag/
├── dag.py           # Node, DAG, DAGBuilder
├── cache.py         # Content-addressed file cache
├── executor.py      # Base executor + registry
├── engine.py        # DAG execution engine
├── activitypub/     # Identity + ownership
│   ├── actor.py     # Actor identity with RSA keys
│   ├── activity.py  # Create, Announce activities
│   ├── signatures.py # RSA signing/verification
│   └── ownership.py # Links actors to assets
├── nodes/           # Built-in executors
│   ├── source.py    # SOURCE
│   ├── transform.py # SEGMENT, RESIZE, TRANSFORM
│   ├── compose.py   # SEQUENCE, LAYER, MUX, BLEND
│   └── effect.py    # EFFECT (identity, etc.)
└── effects/         # Effect implementations
    └── identity/    # The foundational identity effect

Registry: https://git.rose-ash.com/art-dag/registry - Asset registry with ownership proofs
Recipes: https://git.rose-ash.com/art-dag/recipes - DAG recipes using effects

License

MIT

3.1 KiB Raw Blame History