AhmadHassan-BTed/Jantt
AI-NATIVE JSON GANTT ENGINE
Declarative JSON Gantt chart engine. Runs in browser, React, Node.js CLI and autonomous AI agents with built-in DAG solver and critical path analysis.
// KEY FEATURES
- Pure TypeScript, single dependency (fflate), ~63 KB gzipped JS + 8 KB CSS
- Declarative JSON schema for LLMs: 10x fewer tokens than JSX
- Built-in CPM, EVM (ANSI/EIA-748), DCMA-14 audit and CRDT collaboration
- 5 zoom scales, 90° orthogonal links, drag-and-drop, Swiss Noir theme
// DEEP DIVE: HOW IT WORKS
Jantt renders interactive Gantt charts from declarative JSON: drag-and-drop, 5 zoom levels, local storage. Core @jantt/core is pure TypeScript with one dependency (fflate), ~63 KB gzipped JS + 8 KB CSS. Works via script tag, in React, Next.js, Node.js CLI, and as a tool for autonomous AI agents.
Key difference — declarative JSON schema instead of imperative API (like DHTMLX, Frappe) and static DSL (Mermaid). This lets LLMs generate valid timelines with 10× fewer tokens. Inside: topological DAG solver, critical path method (CPM), EVM (ANSI/EIA-748), DCMA-14 audit, CRDT collaboration with roles. Theming via CSS variables: Swiss Light/Dark, Cyber Emerald, Midnight Rose, Sunset Crimson.
Requires modern browser with ES modules and CSS variables; Node.js for CLI. No built-in backend — CRDT sync via cloud rooms. Offline only localStorage, no PM system integrations, no commercial support declared.
// USE CASES
Best for: For AI agents, engineers and PMs to plan and visualize timelines
- Generate valid JSON timelines with LLM agents without coordinate hallucinations
- Perform CPM/EVM analysis and DCMA-14 schedule audit in browser or CLI
- Collaboratively edit charts in real-time via CRDT rooms with roles
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram