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.

2.12x

Quality

—

Does it follow best practices?

Impact

100%

2.12x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Quality

Content

92%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is lean, highly actionable, and workflow-clear, with concrete commands and a sensible iteration loop. Its main weakness is progressive disclosure: everything lives in one file with no references to deeper material.

Suggestions

Move the worked examples (Complete Design Flow, Quick Mobile Mockup) and the prompt-engineering pattern library into a references/ file (e.g. EXAMPLES.md) and link to it from a concise overview, so SKILL.md stays a lean entry point.

De-duplicate the model-selection info: the modelId details appear both in the 'Generate a Screen' parameters and in the standalone 'Model Selection' section — keep one authoritative spot and cross-reference it.

Add an explicit verification checkpoint after generation (e.g. 'call list_screens/get_screen to confirm the screen exists before fetching code') to make the iteration loop's validation step explicit.

Dimension	Reasoning	Score
Conciseness	The body is dense and table-driven with no explanatory padding of concepts Claude already knows (no 'what is Tailwind' or 'what is a UI' prose); nearly every token earns its place, though the model-selection info is restated once.	3 / 3
Actionability	Provides fully executable, copy-paste-ready `mcporter call stitch.*` commands with all parameters, complete end-to-end examples (Complete Design Flow, Quick Mobile Mockup), and an error→fix table.	3 / 3
Workflow Clarity	The 'Iteration Workflow' gives a clear 5-step sequence (create project → generate with FLASH → refine → generate with PRO → extract code) with a feedback loop (present output_components suggestions, regenerate if accepted) and error-recovery guidance; the destructive/batch cap does not apply to generative work.	3 / 3
Progressive Disclosure	Well-organized into clear sections, but it is a single monolithic ~220-line file with no bundle references and content (examples, prompt-engineering patterns) that could be externalized; it does not meet the score-3 bar of well-signaled one-level-deep references and appropriately split content.	2 / 3
	Total	11 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concise, concrete, and complete: it names the tool, enumerates specific actions, and provides an explicit 'Use when' trigger clause with natural terms. It carves a clear, distinct niche with minimal conflict risk.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Generate UI designs and frontend code', 'create screens, mockups, UI designs', 'generate frontend code from text descriptions' — matching the score-3 anchor rather than the domain-only score-2 example.	3 / 3
Completeness	Explicitly answers both what ('Generate UI designs and frontend code with Google Stitch via MCP') and when ('Use when asked to create screens, mockups, UI designs...') with a clear 'Use when' trigger clause.	3 / 3
Trigger Term Quality	Natural user phrasing is well covered — 'create screens', 'mockups', 'UI designs', 'frontend code', 'text descriptions' — terms a user would actually say, not just technical jargon.	3 / 3
Distinctiveness Conflict Risk	Tied to a specific tool (Google Stitch via MCP) and a distinct output niche (screens/mockups/UI designs), making it unlikely to trigger for unrelated skills; 'generate frontend code' has minor overlap with general code-gen skills but the primary niche is clear.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: Dicklesworthstone/pi_agent_rust
Commit: abece4a

Reviewed: 4 days ago

Table of Contents

Discovery Implementation Validation

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.