
housestyle
io.github.myanptlPublished by GitHub account myanptlv1.0.2Updated Oct 24, 2026
Audits CSS for the generic AI-built look, then fixes it: fonts, contrast, palettes, type scales.
Overview
A local MCP server that audits CSS for generic AI-generated design tells and suggests fonts, palettes, type scales and contrast fixes.
- What you can do
- Audit a stylesheet, token block or Tailwind config for ten deterministic generic-design tells, with reasons and fixes for each finding (audit_css). Suggest fonts and font pairings ranked for a role and mood, with ready-to-paste imports and CSS custom properties (suggest_fonts, suggest_pairing). Check colour contrast with WCAG 2.2 AA/AAA verdicts plus an APCA Lc value (check_contrast). Build an 11-step OKLCH colour ramp from a seed colour, with contrast against white and black per step (build_palette). Build a fluid modular type scale emitted as clamp() (build_type_scale).
- When to use it
- Useful when an assistant is generating or reviewing front-end styling and the result risks looking like default AI output: framework blues, Inter, gradients, uniform radii, flat type scales. Good for tightening typography, palette and contrast decisions before shipping a UI.
- Requirements
- Runs locally over stdio, typically started with npx housestyle-mcp; Node.js is needed. No API key, account or configuration. Network access is used only for the Google Fonts public metadata endpoint, cached for an hour; all other tools are pure computation.
Installation
In SourceWeft
- Open housestyle in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
housestyle-mcp
Spots CSS that looks AI-made and suggests fixes for fonts, colour and contrast.
MIT licence.
Source: README.md at commit 0500a30
Tools
0Version history
1- v1.0.2LatestSep 24, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts