Compact symbolic notation for Grok-powered agentic systems
Aglyphs lets you describe complex AI agent architectures, teams, memory systems, and workflows in one compact line — perfect for docs, code comments, whiteboards, and GitHub READMEs.
⨁(𝔾) ⇑ {𝔸_research[𝕄ᵛ], 𝔸_code[𝕋]} → ▷Made for builders who hate bloated diagrams.
- Single Grok ReAct Agent:
𝔾[Λˣ ⊗ 𝕋] (⊙ → Λ → ⊚ → ▷) ⟳ - Grok as Operator:
⨁(𝔾) [Λˣ + 𝕄ˡ + 𝕋] ⇑ {team} - Parallel Team:
⨁(𝔾) (𝔸₁ ⊕ 𝔸₂) → ▷ - With Human Oversight:
𝔾 → ⨁ → ⊕ → ▷
See AGLYPHS-NOTATION.md for the full specification and pattern library.
- Ultra-compact — describe entire teams in < 60 chars
- Text-native — works everywhere (no images needed)
- AI-native — symbols for reflection, memory types, orchestration
- Extensible — easy to add your own glyphs
- Git-friendly — pure text, diffs beautifully
Just copy the glyphs into your Markdown. For better rendering, use a font that supports mathematical symbols (most modern ones do).
Recommended fonts:
- JetBrains Mono
- Fira Code
- DejaVu Sans Mono
- Any Unicode-friendly editor font
For a live playground, see the web demo (coming soon).
- Version: 0.2.0 (2026-03-25)
- License: MIT
- Status: Early but stable — open to community patterns and renderer contributions
AGLYPHS-NOTATION.md— Complete glyph reference, syntax, and ready-to-use patternsCOMPARISON.md— Comparison with Mermaid, LangGraph, CrewAI, and formal notations
We welcome contributions from both developers and autonomous agents.
High-value contributions:
- New patterns for emerging paradigms (Mixture-of-Experts, multi-agent debate, critic loops, etc.)
- Additional glyphs with strong justification
- Font embedding / CSS for better rendering
- JavaScript/Mermaid renderer or VS Code extension
- Real-world usage examples from production agent systems
See CONTRIBUTING.md for guidelines.
Made for the agentic future.
— Brandon Thompson & Blink (2026)