Skip to main content
AI/MLjeremylongshore

engineer-design-diagram

"Generate production-grade engineering design diagrams (architecture,\

Stars
2,267
Source
jeremylongshore/claude-code-plugins-plus-skills
Updated
2026-05-31
Slug
jeremylongshore--claude-code-plugins-plus-skills--engineer-design-diagram
View on GitHubRaw SKILL.md

// install — copy + paste into any project

mkdir -p .claude/skills && curl -fsSL https://raw.githubusercontent.com/jeremylongshore/claude-code-plugins-plus-skills/HEAD/plugins/devops/engineer-design-diagram/skills/engineer-design-diagram/SKILL.md -o .claude/skills/engineer-design-diagram.md

Drops the SKILL.md into .claude/skills/engineer-design-diagram.md. Works with Claude Code, Cursor, and any agent that loads SKILL.md files from .claude/skills/.

Engineer Design Diagram

Generates production-grade engineering design diagrams as single-file HTML with inline SVG, grounded in real repository topology. Credit: design palette + arrow-masking pattern inspired by Cocoon AI's architecture-diagram-generator (MIT). See THIRD_PARTY_LICENSES.md.

Overview

Most diagramming tools produce pretty pictures disconnected from reality. This skill does the opposite: it reads the actual repo (package manifests, docker-compose, k8s, terraform, import graph) and emits a diagram that reflects the real system. It also knows how the system changed — PR-diff mode highlights structural deltas, trace mode turns a stack/log into a sequence diagram, and drift mode detects when the architecture has wandered from a stored fingerprint.

Four modes share a common pipeline (DCI grounding → node/edge graph → template fill → fingerprint write). Output is a single self-contained HTML file that opens in any browser, plus a Mermaid text block for copy-paste into docs. Dark theme with semantic OKLCH color coding by component role. Accessible by default (ARIA, <title>/<desc>, reduced-motion, keyboard navigation).

Layout Philosophy — pick the shape before you draw

Dense technical systems want to render wide. This is how Anthropic docs, Linear docs, and Vercel architecture pages present multi-component systems: a sticky left rail for context (nav, invariants, legend) and a generous main column for the diagram itself. Mimic that pattern when your content is dense, and use the simpler single-SVG hero when it isn't.

Two supported output shapes, one decision up front:

Shape Use when Template
Single-SVG hero (classic) ≤8 nodes, one-screen takeaway, no sub-grouping, no accompanying explanation needed templates/base.html
Docs-layout page (widescreen) ≥8 nodes, multiple planes/groupings/sub-blocks, want invariants + detail cards + legend alongside the diagram templates/docs-layout.html

Widescreen-first for docs-layout. Target min-width: 1024px; do not add mobile breakpoints for docs-layout output — it's architecture documentation, not a landing page. The diagram needs horizontal breathing room. On narrow viewports the diagram stage scrolls horizontally inside its card while the sidebar stays visible.

Hybrid rendering in docs-layout mode: HTML/CSS for all node cards (easier to maintain, free hover states, content edits don't trigger collision math), SVG overlay positioned absolutely over the node grid for arrows only. Arrows are the one place SVG still wins — fixed pixel coords, clean arrowheads, no brittle CSS-line math. Pure-SVG stays the default for the single-hero shape because the payoff of HTML flex doesn't materialize at small node counts.

Layout-decision signal: if during Step 2 the graph has any of — (a) more than one semantic lane/plane, (b) subcomponent lists of 4+ items per node, (c) the user asks for "docs page" / "architecture page" / references Anthropic/Linear/Vercel docs as exemplars — pick docs-layout. Otherwise stay with single-SVG.

See docs-layout.md for the full widescreen spec (grid geometry, design tokens, sidebar anatomy, arrow overlay placement, hover states).

Prerequisites

  • Git repository (for generate/diff/watch modes)
  • At least one of: package.json, pyproject.toml, Cargo.toml, go.mod, docker-compose.yml, k8s/*.yaml, terraform/*.tf
  • Python 3.9+ (for scripts/fingerprint.py)
  • A browser to view output
  • Optional: kubectl (cluster introspection), terraform (state parsing), jq (manifest parsing)

Environment Detection (DCI)

!git rev-parse --show-toplevel 2>/dev/null || echo "not a git repo" !git rev-parse --short HEAD 2>/dev/null || echo "no-head"; git branch --show-current 2>/dev/null || echo "no-branch" !ls package.json pyproject.toml Cargo.toml go.mod requirements.txt docker-compose.yml Dockerfile 2>/dev/null | head -10 || echo "no manifests" !command -v jq >/dev/null 2>&1 && jq -r '.name,(.dependencies // {} | keys[]?)' package.json 2>/dev/null | head -20 || echo "no package.json / no jq" !docker compose config --services 2>/dev/null | head -30 || echo "no docker-compose" !find . -maxdepth 3 -type d \( -name k8s -o -name kubernetes -o -name manifests \) 2>/dev/null | head -5 || echo "no k8s dir" !find . -maxdepth 3 -name '*.tf' 2>/dev/null | head -10 || echo "no terraform"

Instructions

Step 1: Mode Detection

Parse the first token of $ARGUMENTS:

Invocation Mode Playbook
/design:generate or empty generate mode-playbooks.md#generate
/design:diff diff mode-playbooks.md#diff
/design:trace <path> trace mode-playbooks.md#trace
/design:watch watch mode-playbooks.md#watch

All modes share Steps 2-5. Mode-specific variations documented in the per-mode playbook.

Step 2: Build the Node/Edge Graph

Use the DCI block output + targeted reads to populate a working graph:

  • Nodes from: docker-compose service names, k8s kind: Deployment/Service/StatefulSet, terraform resource blocks, package manifest names (for standalone apps), detected binaries in bin/ or cmd/.
  • Edges from: docker-compose depends_on + exposed ports, k8s Service selectors + NetworkPolicy, terraform resource references, import-graph via Grep (import.*from, require(, use crate::), exposed HTTP/gRPC routes.
  • Role classification per node using drawing-rules.md § Role inference. Defaults to slate (external/unknown) when heuristics don't match.
  • Groups from: k8s namespaces, terraform modules, docker-compose networks, directory boundaries for monorepos.

For graphs with >50 nodes, fall back to Mermaid (see Step 3 template selection).

Step 3: Choose Template and Render

Apply the layout-selection rule from Layout Philosophy first, then pick the template for the chosen mode:

Mode Layout shape Template Output
generate / diff single-SVG hero (simple graphs) templates/base.html Centered SVG on dark canvas, cards row below
generate / diff docs-layout page (dense graphs) templates/docs-layout.html Two-column widescreen page: sticky sidebar + main column with diagram card + detail grid
diff (any shape) — selected template + delta classes delta-added / delta-removed / delta-changed CSS markers on changed nodes/edges
trace sequence templates/sequence.html Sequence diagram with lifelines and message arrows
watch — none — markdown drift report Markdown only; no HTML render

For single-SVG hero: fill placeholders per drawing-rules.md — color palette, SVG arrow-masking, z-order, 40px spacing, dashed boundaries, legend placement. Keep the grid <pattern>, #020617 canvas, pulsing header dot (with reduced-motion guard).

For docs-layout page: fill placeholders per docs-layout.md — GitHub-inspired palette (#0f1117 / #161b22 / #1c2128), Inter + JetBrains Mono, 260px sticky sidebar, fixed-pixel node stage with SVG arrow overlay, hover states on node cards. No mobile breakpoints. Populate the sidebar with: (a) version badge, (b) "On this page" nav, (c) "Architectural invariants" list (surface up to 5 load-bearing constraints you inferred from the code/config — e.g. "Kernel owns durable state", "All ops return Result<T,E>"), (d) node-type legend.

If node count >50 OR the model signals layout failure (overlapping boxes, arrows crossing through nodes), switch to templates/mermaid-fallback.html regardless of layout shape.

Step 4: Update Fingerprint State

For generate/diff/watch modes, write structural state to ${CLAUDE_PLUGIN_DATA}/arch-state.json (or ~/.claude-state/arch-state.json fallback if the env var is unset):

python3 ${CLAUDE_SKILL_DIR}/scripts/fingerprint.py write --input /tmp/graph.json

Schema documented in fingerprint-spec.md. Fingerprint persists across sessions so watch mode can detect drift.

Step 5: Validate and Open

Run the HTML validator before presenting output:

python3 ${CLAUDE_SKILL_DIR}/scripts/validate_html.py $CWD/.arch/<mode>-<timestamp>.html

Validator confirms: ARIA labels present, reduced-motion rule exists, no unexpected external script sources beyond Google Fonts. If validation fails, iterate on the template fill — don't ship an inaccessible diagram.

Open the result via ${CLAUDE_SKILL_DIR}/scripts/open_in_browser.sh (OS-aware: xdg-open on Linux, open on macOS, wslview on WSL). Echo the Mermaid equivalent to the chat as a copy-pasteable text block.

Feedback Loop

  1. Run validate_html.py after Step 5.
  2. If ARIA or contrast checks fail → fix template fill → re-validate.
  3. If >50 nodes rendered as overlapping boxes → restart Step 3 with Mermaid fallback.
  4. (docs-layout only) If arrows cross non-endpoint nodes OR text overflows a node box → re-run layout math (see docs-layout.md § verification) before re-screenshotting. Don't eyeball it; collision math is cheap and catches what a thumbnail hides.
  5. Maximum 3 iterations before falling back to Mermaid and flagging for manual review.

Output

  • generate / diff / trace modes: one HTML file at $CWD/.arch/<mode>-<timestamp>.html (self-contained, offline-capable, Google Fonts as the sole external dep) + one Mermaid text block echoed to chat.
  • watch mode: markdown drift report with sections Added, Removed, Changed, each line citing the source file that justified the delta.
  • All modes: updated fingerprint at ${CLAUDE_PLUGIN_DATA}/arch-state.json.

Detailed output contracts in mode-playbooks.md.

Examples

Example 1 — Generate architecture view from a monorepo

Input:

/design:generate

Behavior: DCI auto-loads docker-compose.yml services (web, api, db, cache), classifies roles (frontend/backend/db/db), reads web/src/lib/api-client.ts for the HTTP edge to api, reads docker-compose.yml depends_on for api→db and api→cache edges. Fills templates/base.html, writes ~/.arch/generate-2026-04-19T12-00.html, writes fingerprint with 4 nodes + 3 edges.

Output excerpt (Mermaid block):

flowchart TB
  web["Web Frontend"]:::frontend --> api["API Service"]:::backend
  api --> db["Postgres"]:::db
  api --> cache["Redis"]:::db

Example 2 — PR delta on a branch that adds Redis

Input:

/design:diff

Behavior: Loads prior fingerprint, re-runs DCI on working tree, computes set-diff. Renders architecture view with class="delta-added" on the cache node and the api→cache edge. Summary line: Added 1 node (cache), 1 edge (api→cache).

Example 3 — Sequence diagram from a Sentry event

Input:

/design:trace ./incidents/sentry-2026-04-18-payment-timeout.json

Behavior: Parses Sentry JSON (exception frames + transaction spans), infers 4 lifelines (checkout-api, payment-service, stripe-webhook-listener, fraud-check), renders sequence diagram with the timeout marked on the final arrow. No fingerprint write in trace mode.

Example 4 — Drift watch after a quiet sprint

Input:

/design:watch

Behavior: Loads prior fingerprint, re-runs DCI, diffs. Output:

## Drift report — 2026-04-19 vs 2026-04-12

### Added
- Node `recommendations-service` (backend) — source: docker-compose.yml:services.recommendations
- Edge `api` → `recommendations-service` — source: api/src/handlers/product.ts:44

### Removed
- Edge `api` → `legacy-search` — source: api/src/handlers/search.ts removed in f8a2c91

Edge Cases

  • No package manifests: DCI returns empty. Ask the user to describe the system in prose, fall back to prompt-only generation without DCI grounding.
  • >150 components: Hard cap at 150 rendered nodes. Overflow bucket labeled … +N more; full list preserved in fingerprint.
  • Stack trace format unknown: Three parsers (Sentry JSON, OTel span, raw text). Unknown formats emit Mermaid-only sequence rather than failing.
  • No main branch for diff: Probe origin/HEAD → main → master → trunk. If none found, fall back to HEAD~1.
  • Docker compose binary naming: Probe both docker compose (modern) and docker-compose (legacy).
  • ${CLAUDE_PLUGIN_DATA} unset: Fallback chain ${CLAUDE_PLUGIN_DATA} → ${XDG_STATE_HOME}/claude/arch → ~/.claude-state/arch.
  • iOS Safari large-diagram rendering: Avoid SVG paths with >10k points per element; fall back to Mermaid for graphs that approach this limit.
  • k8s manifests with CRDs: Role defaults to slate (external/unknown) when no heuristic matches; document in the legend.

Error Handling

Error Cause Solution
not a git repo Running outside git Init a repo or cd to one; DCI handles the empty case but diff/watch require git
no manifests No package files found Use prose-only mode; skill will ask user to describe the system
fingerprint schema mismatch State file from older schema version fingerprint.py refuses cross-version diff and emits baseline-restart guidance
ARIA validation failed Template fill omitted <title>/<desc> Re-fill with role-appropriate labels from accessibility.md
overlapping boxes detected Too many nodes for SVG layout Auto-switch to Mermaid fallback (node count >50 trigger)
no diff base main/master/trunk all missing Use HEAD~1 as base, warn user in output
kubectl not installed Live cluster introspection unavailable Skip cluster overlay; diagram still renders from static manifests
trace parser unknown format Input file matches no known schema Emit Mermaid-only sequence with best-effort parsing, flag for user review

Resources