Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →

googlechrome/modern-web-guidance-src

8 skills16 installs7.4k starsGitHub

Install

npx skills add https://github.com/googlechrome/modern-web-guidance-src

Skills in this repo

1Project GuidesDefines best practices for writing self-contained guide.md files that are the only source an AI coding agent sees for a web-feature use case. A developer uses it when authoring or reviewing web-platform implementation guidance.3installs2Baseline StatusVerifies whether a web feature is ready to use by checking its browser support and Baseline status through a CLI query. A developer uses it to confirm cross-browser readiness of a feature.2installs3Investigating Eval ResultsDiagnoses why AI coding agents fail guidance evaluations by comparing guided vs unguided runs and debugging grader failures. A developer uses it to investigate low guided pass rates and fix task, grader, or expectation files toward a 100% pass rate.2installs4Project Discipline GuidesRefactors discipline-level web guides such as JavaScript and CSS to strip out common knowledge using generated model-specific knowledge mirrors. A developer uses it to keep guidance focused on non-obvious content.2installs5Project EvalsProvides best practices for authoring expectations and grader files that measure guidance quality. A developer uses it when writing or reviewing an expectations.md or grader.ts file.2installs6Project Guide ValidationDefines a protocol for validating the technical accuracy, framework nuances, and eval-readiness of web guidance. A developer uses it when assigned to validate or review a guide, demo, or expectations file.2installs7Project Use CasesProvides best practices for writing use cases for a given web feature as the first step in creating a new guide. A developer uses it when authoring or reviewing a use case under the guides directory.2installs8Nightly Eval InvestigationRetrieves recent nightly evaluation runs from Google Cloud Storage and compares them across three agents to flag unhealthy, brittle, or over-prescribed guides. A developer uses it for a read-only bulk investigation of remote nightly eval health.1installs

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.

googlechrome/modern-web-guidance-src · 8 skills · Skillselion