HyperWeave is an artifact layer that enables AI agents to generate verifiable, portable, and self-describing visual outputs like diagrams and matrices. It bridges the gap between raw agent output and human understanding by embedding machine-readable metadata directly into SVG artifacts.

Is hyperweave open source?

Yes, licensed under Apache-2.0.

hyperweave: Artifact layer that enables AI agents to…

Features

01Self-Describing Artifacts: Generates SVG artifacts that embed their complete specification and a hash-verified digest, allowing agents to re-ingest, verify, and modify them without parsing pixels.

02Diverse Diagram Types: Supports various diagram topologies like pipeline, fanout, DAG, state-machine, and tree, rendering them as self-contained SVGs with embedded specs.

03Structured Matrices: Creates structured tables rendered as portable SVGs, including comparison grids, benchmarks, and heatmaps, with machine-readable payloads for agents.

04Agentic Session Receipts: Automatically parses agent coding sessions (e.g., Claude Code, Codex) into portable SVG receipts detailing model usage, token spend, and context load.

05Aesthetic Genomes (Themes): Provides configurable 'genomes' (themes) for consistent visual identity across all artifact types, allowing agents to apply predefined aesthetic specifications.

Compatibility

Python

Runtime

Verified via docs

AI Agents

Agent Output

Verified via docs

SVG

Output Format

Verified via docs

CLI

Interface

Verified via docs

HTTP API

Interface

Verified via docs

Quick start

$ uv add hyperweave

Use cases

↳Visualizing Agent Plans and Outputs: Agents can generate diagrams, tables, and receipts to show their work, plans, and explanations in a human-readable and verifiable format.

↳Embedding Explanations in Documents: Portable SVG artifacts can be embedded into GitHub READMEs, Slack messages, emails, or other documents, retaining their structure and metadata.

↳Agent-to-Agent Communication: Artifacts serve as a common, verifiable representation for agents to exchange structured information, enabling easier inspection, transformation, and regeneration.

↳Benchmarking and Comparison: Creating structured matrices for comparing LLM models, project metrics, or feature sets, automatically rendered from structured data.

hyperweave

Features

Compatibility

Quick start

Use cases

Alternatives

Comments

Deploy on DigitalOcean