Wikimedia
Official@wikimedia · United States of America
Wikimedia is a global movement whose mission is to bring free educational content to the world.
Agent Skills by Wikimedia
Showing 25 vetted skills indexed across 2 GitHub repositories.
wikimediacodex-style-guide
Apply Codex design principles to UX decisions for color, typography, and layout.
wikimediaprotowiki-update-codex
Upgrade ProtoWiki Codex packages in lockstep and propagate documented changes.
wikimediawiki-apis
Guide REST vs Action API selection for Wikimedia data tasks.
wikimediacodex-components
Catalog Codex Vue components and link to their reference documentation.
wikimediacodex-typography
Enforce canonical text styles across prototype UI typography.
wikimediaprotowiki-components
Catalog ProtoWiki's shipped UI components and their documentation locations.
wikimediacodex-usage
Adopt Wikimedia Codex design system components, tokens, and icons in frontend projects.
wikimediaprotowiki-theme
Resolve ProtoWiki global theming into a data-theme attribute at boot.
wikimediaprotowiki-deploy
Automate ProtoWiki deployment to GitHub Pages with PR previews and base-path handling.
wikimediaprotowiki-getting-started
Summarize ProtoWiki's purpose, tech stack, and repository structure for onboarding.
wikimediaprotowiki-create-prototype
Create prototype folders with index.vue under src/prototypes/ for automatic gallery publishing.
wikimediaprotowiki-snapshot-data
Coordinate ProtoWiki snapshot workflows across fetchers, HTML snapshots, and skin scoping.
wikimediawiki-snapshot-data
Snapshot Wikipedia article HTML and skin CSS into offline fixtures.
wikimediacodex-icons
Import constants from @wikimedia/codex-icons and render them with CdxIcon in Vue templates.
wikimediaprotowiki-skins
Resolve desktop or mobile skins at boot via URL parameter or viewport.
wikimediacodex-tokens
Standardize Codex design tokens across ProtoWiki themes and components.
wikimediawiki-signals
Catalog Wikipedia data streams from Wikimedia REST and analytics endpoints.
wikimediaoptimize-model
Identifies CPU/GPU bottlenecks in KServe-hosted Python models and outputs prioritized fix diffs.
wikimediabump-image
Update LiftWing Docker image tags in Helm values.yaml files.
wikimediaincident-triage
Correlate Prometheus signals, deployment changes, and task metadata to diagnose LiftWing incidents.
wikimedialocal-test
Validate Lift Wing KServe-style model servers locally with Docker Compose and curl prediction requests.
wikimediaadd-ml-service
Scaffold and register LiftWing inference services in KServe helmfile and values manifests.
wikimediatroubleshoot-ml-k8s
Diagnose failing Wikimedia KServe/Knative InferenceService deployments using kubectl evidence.
wikimediaadd-model
Scaffold KServe inference service directories with Docker Compose, Blubber configs, pipelines, and CI wiring.
Frequently Asked Questions About Wikimedia
FAQPage SchemaWhat specific tasks can I perform using these design and infrastructure capabilities?▼
You can standardize frontend interfaces using Codex components and tokens, scaffold new inference services for KServe, manage container image versions in Helm manifests, and perform structured code reviews for Gerrit-based repositories.
Which technical personas are these capabilities designed for?▼
These capabilities are intended for frontend engineers building interfaces with the Codex system, and site reliability or machine learning engineers managing model serving infrastructure on Kubernetes clusters.
What are the prerequisites for deploying a new inference service?▼
To deploy a new service, you must have a configured KServe environment, access to the LiftWing infrastructure, and the ability to define service directories containing Blubber configurations and Helm values manifests.