Skills for Real Engineers. Straight from my .agents directory.
excalidraw
Agent skill fromsoftaworks/agent-toolkit
Use when working with *.excalidraw or *.excalidraw.json files, user mentions diagrams/flowcharts, or requests architecture visualization - delegates all Excalidraw operations to subagents to prevent context exhaustion from verbose JSON (single files: 4k-22k tokens, can exceed read limits)
How to install excalidraw
npx skills add softaworks/agent-toolkit# or add only this skill to Claude Code: copy the folder
skills/excalidraw/
# into
~/.claude/skills/excalidraw/ (all projects)
.claude/skills/excalidraw/ (one project)
1# Excalidraw Subagent Delegation23## Overview45**Core principle:** Main agents NEVER read Excalidraw files directly. Always delegate to subagents to isolate context consumption.67Excalidraw files are JSON with high token cost but low information density. Single files range from 4k-22k tokens (largest can exceed read tool limits). Reading multiple diagrams quickly exhausts context budget (7 files = 67k tokens = 33% of budget).89## The Problem1011Excalidraw JSON structure:12- Each shape has 20+ properties (x, y, width, height, strokeColor, seed, version, etc.)13- Most properties are visual metadata (positioning, styling, roughness)14- Actual content: text labels and element relationships (<10% of file)15- **Signal-to-noise ratio is extremely low**1617Example: 14-element diagram = 596 lines, 16K, ~4k tokens. 79-element diagram = 2,916 lines, 88K, ~22k tokens (exceeds read limit).1819## When to Use2021**Trigger on ANY of these:**22- File path contains `.excalidraw` or `.excalidraw.json`23- User requests: "explain/update/create diagram", "show architecture", "visualize flow"24- User mentions: "flowchart", "architecture diagram", "Excalidraw file"25- Architecture/design documentation tasks involving visual artifacts2627**Use delegation even for:**28- "Small" files (smallest is 4k tokens - still significant)29- "Quick checks" (checking component names still loads full JSON)30- Single file operations (isolation prevents context pollution)31…
Excerpt shown under the repository's MIT license.
More skills in agent-toolkit
Similar skill repositories
Public repository for Agent Skills
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Production-grade engineering skills for AI coding agents.
The 1st Collaborative AI Design Agent Workspace. Local-first and open source. Create motion, prototypes, websites, apps, dashboards, presentations, images, and videos with your team and AI agents. Share work, exchange feedback, and iterate together in one workspace.
Turn any idea, plan, or codebase into a beautiful interactive diagram. An agent skill for Claude Code, Codex, and more.
