Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/onexeor-dev/lumo/skillnpx skills add OneXeor-Dev/lumo --skill skillgit clone --depth 1 https://github.com/OneXeor-Dev/lumoWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00104 | $0.11976 |
| Opus 5 | $0.00052 | $0.05988 |
| Sonnet 5 | $0.00021 | $0.02395 |
| Haiku 4.5 | $0.00010 | $0.01198 |
Grade A, and why
lumo scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 1,017 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lumo
Mobile design intelligence for Claude Code. Lumo combines a curated rule set with a small toolkit of deterministic Python tools — so checks that need real math (WCAG luminance, OKLCH correction, AST diff, cross-platform parity) return facts, not LLM guesses.
When to Use
Use Lumo when the user request touches any of:
- Accessibility / WCAG contrast questions on a colour pair or palette
- Reviewing or auditing mobile UI code (Jetpack Compose, Android XML, SwiftUI, UIKit)
- Comparing the same screen across iOS and Android for visual / behavioural parity (padding, animation timing, touch target, typography scale)
- Applying cognitive-science rules to a layout: Fitts (target size × distance), Hick (choice count → decision time), Gestalt grouping, Nielsen heuristics
- Choosing or validating colour tokens for a mobile design system
- Detecting platform-specific anti-patterns (e.g. emoji as system icons, hardcoded hex in components, missing safe-area handling)
When NOT to Use
Skip Lumo entirely for:
- Pure backend / API / database tasks
- Non-mobile web design (use a web-focused skill instead)
- Build / CI / dependency / version-bump tasks
- Performance work that doesn't touch UI rendering
- Tasks where the user has explicitly opted out of design feedback
Do not invoke Lumo "just in case" the request might be design-adjacent. If the request doesn't match a trigger above and the user hasn't named a mobile UI concern, stay out.
Tools
Lumo ships with Python tools. Each tool has a single CLI entry-point with
deterministic output. Invoke via Bash. Tools live in tools/lumo/<area>/cli.py
and are exposed as console scripts after pip install -e tools/ or
pipx install lumo-mobile.
When running inside an MCP-aware client (Cursor, Continue, Aider, Goose,
Zed, Codex, or Claude Code with MCP enabled), the same tools are
also exposed as MCP functions: lumo_wcag_check, lumo_wcag_fix,
lumo_theory_check, lumo_parity_diff, lumo_source_check_compose,
lumo_source_check_swiftui, lumo_audit_scan, lumo_figma_diff,
lumo_figma_render, lumo_render_compose, lumo_render_swiftui.
Prefer the MCP function over spawning a Bash subprocess when available
— the structured response is already JSON and the user does not see
noisy command output.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 1,017 lines · 104 tokens per session scan A 604b0a705c33
lumo is a skill published in the GitHub repository OneXeor-Dev/lumo (2 stars, last pushed 2mo ago), licensed MIT. It adds 104 tokens to every session and 11,976 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
driftx
Visual comparison, accessibility audit, layout regression, and device interaction for mobile apps (Android, iOS, React Native). Use when the user asks to compare their app against a design, check accessibility, detect visual regressions, inspect the component tree, capture screenshots, or interact with a…
e2e-testing
AI-powered E2E testing for any app — Flutter, React Native, iOS, Android, Electron, Tauri, KMP, .NET MAUI. Connects via MCP to running apps so the agent can take screenshots, tap elements, enter text, scroll, inspect UI trees, and verify state with natural language. Use when the user wants to test an app's UI…
material-design-3-ui
Design, redesign, review, or implement user interfaces using Google Material Design 3 (M3), Material You, and optional Material 3 Expressive principles. Use for UI/UX screens, app flows, design systems, component selection, theming, responsive/adaptive layouts, accessibility audits, Figma-ready specifications, or…
modal-and-overlay-patterns
Overlays — modals, drawers, bottom sheets, popovers — interrupt or augment the main flow. Each type has a different scope, blocking level, and appropriate use case. Use when designing dialogs, confirmation prompts, side panels, action sheets, or any UI element that appears above the main content layer.
responsive-paradigms
Mobile, tablet, and desktop are different interaction paradigms — not the same layout scaled up or down. Sections can be hidden, repositioned, or made sticky on mobile. Navigation and primary actions move. Use when designing responsive layouts, adapting desktop UI for mobile, or deciding what to show on each…
sticky-and-fixed-elements
Sticky and fixed positioning keeps critical UI persistent as the user scrolls — headers at the top, toolbars at the bottom on mobile. Use deliberately: too many fixed layers create visual noise and reduce content area. Use when designing navigation headers, bottom toolbars, floating action buttons, or table column…