一键导入
ui-audit
Comprehensive workflow to audit the `ambient-ui` component's ui & ux. Use after any changes are made to the `ambient-ui` user interface.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Comprehensive workflow to audit the `ambient-ui` component's ui & ux. Use after any changes are made to the `ambient-ui` user interface.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Install and verify Ambient Code Platform on an OpenShift cluster using quay.io images. Use when deploying Ambient to any OpenShift namespace — production, ephemeral PR test instances, or developer clusters. Covers secrets, kustomize deploy, rollout verification, and troubleshooting.
Deploy a PR's images (api-server, control-plane, runner) to any OpenShift namespace for integration testing. Supports full SSO deployments (install-openshift.sh) and lightweight dev-mode deployments (install-standard.sh).
Log one or more Jira issues to the ENGPROD project with the acp component pre-filled. Gathers context to make issues agent-actionable from cold start. Use this whenever work needs tracking in Jira -- creating stories, filing bugs, logging tasks, opening spikes, creating epics, or any time the user says "create a jira", "log this", "file a bug", "new ticket", "open a story", "track this", "create tickets", "batch create", or provides a bullet list of work items. Supports single tickets and batch creation from bullet lists. Triggers on: "log jira", "create jira", "file a bug", "new ticket", "open a story", "jira issue", "track this work", "open a jira", "create a ticket", "create tickets", "batch tickets", "log these items".
Deploy a Kind cluster with OpenShell gateway mode, apply tenant fleet definitions via acpctl apply -k, set up Vertex/Google credentials, spin up sandbox sessions, and configure local openshell CLI connectivity. Use this skill whenever the user wants to test gateway-provisioned sandboxes, create credentials in tenants, debug sandbox provisioning, connect the openshell CLI to a gateway, or run dual-tenant tests. You SHOULD reach for this skill on: "kind-up", "kind cluster", "openshell gateway", "gateway mode", "OPENSHELL_USE_GATEWAY", "sandbox provisioning", "dual-tenant", "tenant-a", "tenant-b", "vertex credential", "spin up a sandbox", "credential in tenant", "openshell sandbox list", "gateway connectivity", "connect to the gateway", "mtls", "port-forward gateway", "acpctl apply -k", "local development".
Upgrade the rh-trex-ai upstream framework dependency in ambient-api-server. Handles version bump, compilation verification, breaking change detection, test execution, and go.sum cleanup. Use when: "upgrade rh-trex-ai", "bump upstream", "update trex", "new upstream version", "upgrade framework", "pull in upstream changes", "update rh-trex-ai to latest".
Top-level autonomous orchestrator that reconciles all specs against the codebase. Reads skills/RECONCILE.md for checkpoint state, runs field-level gap analysis, plans waves, and executes the full-stack-pipeline per wave. Idempotent: safe to run repeatedly -- produces the same result for the same spec+code state. Use when: "reconcile everything", "run full reconciliation", "what's the gap across all specs", "autonomous build", "just do it", "build everything", "reconcile specs to code", "spec coverage", "what's left to implement", "implementation status".
| name | ui-audit |
| description | Comprehensive workflow to audit the `ambient-ui` component's ui & ux. Use after any changes are made to the `ambient-ui` user interface. |
Follow this workflow precisely and exactly.
Launch 15 subagents with the following personas to perform a comprehensive audit from their perspectives. They should look wholistically at the ui/ux/user journey, but you should point out which changes were recently done so that they can provide specific feedback about new changes, but in the context of the app as a whole.
Each expert should also provide review in-light of the ## Design System section of the specs/ui/index.spec.md - this is a Red Hat tool. All design system items must be adhered to.
Agents should, if possible, use both Playwright and direct code reading to not only understand the structure of the UI, but also the appearance and feel of using the app.
Identify feedback that doesn't require a decision from a human, and apply it directly using subagents. Parallelize where possible to maximize throughput and quality.
Surface any design decisions after kicking off the subagents.
After design decisions have been provided, proceed with implementation.
FOR ANY AND ALL CHANGES, ensure that the relevant UI specs are updated, if needed.
Specs SHALL be the source of truth.
Loop to step 1. Perform the loop 3 times, even if feedback becomes sparse.
Commit changes atomically, and present the result for review.