stitch

Generate UI designs and frontend code with Google Stitch via MCP. Use when asked to create screens, mockups, UI designs, or generate frontend code from text descriptions. Supports desktop, mobile, and tablet layouts.

Install with Tessl CLI

npx tessl i github:Dicklesworthstone/pi_agent_rust --skill stitch

What are skills?

Review — 89%

Does it follow best practices?

If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:

npx tessl skill review --optimize ./path/to/skill

Learn more

Validation — 10 / 11 Passed

Validation for skill structure

SKILL.md

Review

Evals

Evaluation results

100%

47%

Automate UI Prototype-to-Production Design Workflow

Iteration workflow and model selection

Criteria

Without context

With context

FLASH for prototypes

100%

PRO for finals

100%

Project created first

50%

100%

Code extraction step

37%

100%

No retry on timeout

100%

mcporter call syntax

100%

Prompts specify layout

100%

Prompts specify components

100%

Prompts specify style

100%

Multi-screen coverage

100%

Correct model ID spelling

100%

Without context: $0.3865 · 2m 12s · 23 turns · 107 in / 7,331 out tokens

With context: $0.3547 · 1m 25s · 17 turns · 20 in / 4,769 out tokens

100%

49%

Cross-Platform UI Generation for a Productivity App

Device types and prompt engineering

Criteria

Without context

With context

MOBILE deviceType

100%

TABLET deviceType

100%

DESKTOP deviceType

100%

Prompts specify layout

100%

Prompts specify components

100%

Prompts specify content

100%

Prompts specify style

100%

No weak/vague prompts

100%

Design consistency reference

35%

100%

mcporter call syntax

100%

projectId parameter used

100%

Without context: $0.4220 · 1m 38s · 27 turns · 31 in / 5,327 out tokens

With context: $0.8925 · 3m 11s · 43 turns · 50 in / 9,354 out tokens

100%

62%

SaaS Analytics Platform Design Automation

Output suggestions handling and design consistency

Criteria

Without context

With context

output_components checked

100%

Re-call with suggestion

100%

Design consistency reference

100%

fetch_screen_code used

100%

fetch_screen_image used

100%

get_screen for status check

50%

100%

No retry loop

100%

mcporter call syntax

100%

Multiple screens generated

100%

Without context: $0.6471 · 2m 7s · 22 turns · 1,751 in / 7,129 out tokens

With context: $0.5031 · 2m 30s · 18 turns · 178 in / 8,660 out tokens

Evaluated: about 13 hours ago
Agent: Claude Code

Table of Contents

Automate UI Prototype-to-Production Design Workflow Cross-Platform UI Generation for a Productivity App SaaS Analytics Platform Design Automation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.