Browse skills
197 skills — plugin-bundled and standalone — across the catalog. Search by name or filter by platform, discipline, outcome, lifecycle, or tag. New here? See what skills are.
No skills match.
-
Deep ARIA pattern review for components and pages: landmark structure, widget role correctness, live-region design, keyboard interaction model, focus management for modals/drawers/tabs, and state attribute hygiene (aria-expanded, aria-selected, aria-controls). Maps issues to WCAG SC 1.3.1, 2.1.1, 4.1.2, 4.1.3 with concrete markup fixes. Adapts to React, Astro, Next.js, and CMS-rendered output (Sitecore, Optimizely, Umbraco).
-
Recurring WCAG 2.2 AA compliance health-check for live DX properties: compares the current axe-core scan against a saved baseline, surfaces regressions introduced since last audit, tracks sprint-over-sprint improvement, and produces a compliance status dashboard covering ADA, EAA (European Accessibility Act), and Section 508. Designed for managed-services engagements and long-running client programmes.
-
Convert raw WCAG 2.2 AA audit findings into a prioritised, sprint-ready remediation backlog. Groups issues by severity and component, estimates effort, assigns platform owner (design/engineering/content/CMS config), and produces Jira-ready user stories or Azure DevOps work items with acceptance criteria tied to the relevant WCAG SC. Includes a quick-win list for immediate before-sprint fixes and a governance action for content-model / authoring-UX gaps.
-
Draft a VPAT 2.5 Accessibility Conformance Report (ACR) from audit findings. Produces a publication-ready document in VPAT 2.5 WCAG Edition format covering all applicable WCAG 2.2 AA success criteria, with conformance ratings (Supports / Partially Supports / Does Not Support / Not Applicable) and remarks. Suitable for enterprise procurement, Section 508, and European Accessibility Act (EN 301 549) submissions. Adapts output for WCAG, EU (EN 301 549), and INT editions.
-
Full-page WCAG 2.2 AA accessibility scan combining automated axe-core/Playwright checks with AI-judged criteria (alt-text quality, reading order logic, focus sequence, form instruction clarity). Produces a severity-ranked finding list mapped to specific WCAG 2.2 success criteria with pass/fail/partial conformance ratings and immediate remediation pointers. Adapts output for Sitecore XM Cloud, Optimizely CMS, Umbraco, and headless front-ends.
-
Audits Consent Mode v2 wiring end-to-end — default/update sequencing, CMP integration, tag firing guards, IAB TCF/GPP signal handling, and regional configurations — and outputs a compliance gap report with remediation steps. Use before any GDPR/CCPA go-live, after a CMP change, or during a privacy audit.
-
Scans GTM containers, data-layer specs, URL patterns, and event parameter values for Personally Identifiable Information exposure risk — then generates a prioritised redaction task list. Use before any analytics go-live, after a CMS migration, or as part of a privacy audit.
-
Consolidates findings from one or more analytics audit skills (tag audit, consent review, PII scan) into a single prioritised remediation backlog with severity ratings, effort estimates, owner assignments, and sprint-ready ticket descriptions. Use after running any audit to produce a delivery-ready action plan.
-
Audits an existing GTM container and GA4 implementation against a measurement strategy — inventories tags, triggers, and variables; identifies duplicate events, missing parameters, mis-mapped conversions, paused/dead tags, and configuration drift. Use when verifying an implementation before launch, after a CMS migration, or during a recurring managed-services health check.
-
Translates business objectives into a complete GA4 measurement strategy — objectives, KPIs, measurement questions, event taxonomy, data-layer spec, custom dimensions, and governance. Use at the start of any engagement before instrumentation begins, or when re-baselining an existing strategy after a platform migration.
-
Generates 2–3 named architecture options (e.g., "Full MACH Composable", "Hybrid / Incremental Composable", "Modernised Monolith") for a DXP or digital platform engagement, each with a narrative, trade-off summary, indicative complexity, and a recommended default. Use when a client or internal team needs an architecture options paper for a governance review, a discovery deliverable, or a pre-sales response — e.g., "produce architecture options for our re-platforming", "draft an options analysis for headless migration", or "what are our architecture choices?".
-
Translates a selected composable DXP or MACH architecture into a phased delivery roadmap covering discovery, pilot, migration, cutover, and optimise waves. Produces a timeline, workstream breakdown, dependency map, and risk register. Use when a client has agreed on an architecture direction and needs a delivery plan — e.g., "build a roadmap for our Sitecore XM Cloud migration", "create a phased plan for our composable re-platform", or "what does an 18-month DXP migration look like?".
-
Produces a Mermaid architecture diagram of a composable DXP or MACH system, showing the service topology, API boundaries, CDN/edge layer, identity, search, and commerce integrations. Generates C4 Context and Container diagrams alongside a MACH-layer diagram. Use when a project needs a visual architecture artefact for a client presentation, an ADR, a technical design doc, or a governance review — e.g., "draw the architecture for our headless CMS stack", "generate a system integration diagram", or "create a C4 diagram for our composable DXP".
-
Evaluates 2–5 DXP or composable platform candidates against a weighted set of project requirements and produces a red/amber/green scoring matrix with a ranked recommendation and rationale. Use when a client engagement has shortlisted platforms and needs a defensible, stakeholder-ready comparison — e.g., "score Sitecore XM Cloud vs Contentful vs Optimizely for our requirements", "which headless CMS fits our team?", or "create a platform selection scorecard".
-
Builds a 3-year total cost of ownership (TCO) model for a composable DXP or MACH platform architecture, covering licence fees, implementation, integration, infrastructure, operations, training, and a velocity-delta estimate vs. current state. Produces a cost-breakdown table, a net-present-value narrative, and a comparison across 2–3 architecture options. Use when a client needs a business case, a board-level cost summary, or a vendor shortlist with cost dimension — e.g., "estimate the TCO for our Sitecore XM Cloud migration", "compare the 3-year cost of full MACH vs. hybrid composable", or "build a DXP TCO model".
-
Writes nomination-ready award submissions for Sitecore Partner Experience Awards, Optimizely Partner Awards, Webby Awards, and comparable digital experience and agency award bodies. Adapts the challenge/approach/results narrative to each body's word limits, judging criteria, and required evidence sections. Apply when an award window is open and a completed (or near-complete) project is being nominated. Skip when you need a client-facing case study (use case-study-draft) or a partner compliance evidence package (use partner-evidence-package).
-
Drafts a full client-facing case study (challenge / approach / results) from project artifacts and supplied metrics. Produces a web-ready narrative plus a PDF-optimised one-pager variant. Works for any RBA platform engagement — Sitecore, Optimizely, Umbraco, Azure, or composable/headless. Apply when a project has closed (or is in hypercare) and BD or strategy needs a polished case study before the team disperses. Skip when you only need an internal win story (use win-story-brief) or an award submission (use award-submission-writer).
-
Interviews the engagement team and structures raw project data into a reusable metrics register, ready to feed case-study-draft, award-submission-writer, and partner-evidence-package. Reads Azure DevOps exports, Jira exports, Lighthouse reports, and GA4/Adobe Analytics exports when provided as local files. Apply before drafting any case study or award submission when project metrics have not yet been formalised. Skip when solid metrics already exist in the project brief.
-
Assembles partner-tier evidence packages for Microsoft Solutions Partner renewal, Sitecore Gold/Platinum tier, and Optimizely Gold/Platinum tier. Produces structured customer reference documents, deployment proof narratives, capability statements, and evidence checklists aligned to each programme's validation requirements. Apply when a partner tier renewal or specialisation audit is approaching and the delivery team needs to package completed project evidence. Skip when you need a public case study (use case-study-draft) or an award nomination (use award-submission-writer).
-
Produces a concise internal win story (1–2 pages) for sales enablement, new-business pitches, and capability decks. Faster and more candid than a client-facing case study — written for RBA colleagues who need to understand why we won, what we delivered, and how to position it in a proposal. Apply immediately after deal close or project go-live when institutional knowledge is freshest. Skip when you need a client-facing case study (use case-study-draft) or an award submission (use award-submission-writer).
-
Queries Azure DevOps and/or Jira via MCP to pull sprint progress, work-item state, pipeline health, PR throughput, and timesheet burn into a normalised JSON snapshot file ready for status-report-draft. Use at the start of every reporting cycle.
-
Formats a drafted status report markdown file for the target distribution channel — email body, Microsoft Teams adaptive card, Confluence page, or plain markdown — and outputs the ready-to-send payload. Use after status-report-draft produces a report.
-
Reads a status snapshot (from status-data-pull) and drafts a polished status report for the specified audience — executive, client, or engineering. Handles RAG ratings, milestone tables, risk narratives, and next-period previews. Always produces report as a markdown file ready for status-distribute.
-
Generates production-ready Azure Bicep templates for PaaS resource combinations (App Service, Container Apps, SQL, Redis, Key Vault, etc.) with parameter files for each environment tier. Use when a project needs repeatable, reviewer-friendly IaC for an Azure workload or when migrating ARM JSON to Bicep.
-
Produces a comprehensive environment runbook for an Azure deployment — topology diagrams, access patterns, deployment procedures, incident-response steps, and rollback playbooks. Use before go-live on any environment to ensure the platform team can operate it without the original delivery team present.
-
Scaffolds a CAF-aligned Azure landing-zone structure — management-group hierarchy, Azure Policy assignments, hub-spoke networking, Log Analytics, and Defender for Cloud — output as Bicep or Terraform. Use when establishing a new client Azure tenant or maturing an existing subscription towards enterprise governance.
-
Applies the Microsoft cloud security benchmark and RBA's DX workload security standards to an Azure subscription — generating Azure Policy assignments, RBAC role definitions, and a gap-assessment report with prioritised remediations. Use before go-live or during a security review of an Azure DX environment.
-
Generates structured Terraform modules for Azure (AzureRM) or AWS providers — with remote state on Azure Blob Storage or S3, workspace-per-environment parity, and per-environment tfvars. Use when a project uses Terraform as its IaC tool or when converting Bicep/ARM to Terraform.
-
Authors multi-stage Azure Pipelines YAML (or GitHub Actions) for build, test, security scan, and environment-gated deployment — with OIDC service connections, Key Vault variable groups, and approval gates. Use when creating or modernising CI/CD for a DX platform project.
-
WCAG 2.2 AA accessibility review of React/TypeScript, Razor, or HTML markup. Produces per-violation findings with criterion reference (e.g. 1.4.3), impact level (critical/serious/moderate/minor per axe-core taxonomy), and a concrete code fix. Use on any component or template change before QA or client hand-off.
-
Core Web Vitals and server-side performance audit for RBA DX projects. Reviews frontend bundle discipline, image optimisation, render-blocking resources, N+1 query patterns, HTTP caching, and React/Next.js rendering strategy. Produces findings ranked by estimated user impact (LCP / INP / CLS deltas). Use on any frontend component PR or API layer change before client delivery.
-
Full pull-request review against RBA engineering standards: architecture, OWASP Top 10:2025 security, Core Web Vitals performance, WCAG 2.2 AA accessibility, and platform-specific conventions (Sitecore Helix, Optimizely, Umbraco, .NET, TypeScript/React). Use when you want a comprehensive, standards-backed review before a human reviewer sees your PR.
-
Synthesises one or more review outputs (security-scan, perf-audit, a11y-linter, standards-check, pr-reviewer) into a stakeholder-ready PR summary with a clear go/no-go recommendation and a prioritised action list. Use before presenting a PR to the client delivery lead or before a formal client release gate.
-
OWASP Top 10:2025-aligned security scan of a file, diff, or directory. Produces a severity-ranked findings report with CWE references and concrete .NET/TypeScript/platform-specific remediation steps. Use before merging any change that touches auth, data access, APIs, or third-party integrations.
-
Validates code against RBA platform coding standards: Sitecore Helix layer discipline, Optimizely content-type conventions, Umbraco composition patterns, RBA .NET project structure, and TypeScript/React architecture rules. Use when onboarding a new developer to a project, reviewing a platform migration diff, or running a periodic standards audit between sprints.
-
Inventories a source CMS's content types, field schemas, taxonomy, media library, and content relationships, then classifies every item as migrate/rewrite/retire. Use at the start of a migration engagement before any field mapping or technical scoping begins.
-
Strips proprietary markup from exported CMS rich-text content — Word paste artifacts, legacy shortcodes, vendor attributes, inline styles, and platform-specific HTML — and emits clean, portable HTML or plain-text suitable for import into any target CMS. Use before field mapping ETL on any rich-text field.
-
Builds a complete old-URL→new-URL 301 redirect map from a sitemap, crawl export, or content inventory; detects redirect chains, classifies SEO impact, and outputs a deployment-ready redirect ruleset. Use before go-live to prevent organic traffic loss.
-
Validates that all SEO equity — meta titles, descriptions, canonical tags, structured data, hreflang, XML sitemaps, and robots.txt — survives a CMS cutover; produces a scored readiness report with pass/fail for each checkpoint. Use in the final validation sprint before go-live.
-
Produces a source→target CMS field-mapping specification with transformation rules, data-type mismatch notes, and adapter-specific gotchas; use after content-audit to drive the ETL implementation.
-
Maps content types to channels, audiences, personas, and buyer journey stages to produce a prioritised content matrix. Shows which content types serve which audiences across which channels, identifies gaps where personas are underserved, and connects IA decisions to a prioritised implementation backlog. Use when aligning strategy and IA to delivery planning, after content types are defined — e.g. "create a content matrix for our B2B SaaS site", "map our content types to buyer journey stages", or "which content types do we need for our launch MVP vs phase 2".
-
Designs a complete, platform-neutral content type hierarchy from project requirements, existing content audits, or a brief description. Outputs a structured schema with content types, field definitions, field types, inheritance/composition chains, relationship maps, and per-platform implementation notes. Use when starting a new DXP build, migrating from a legacy CMS, or rationalising an inherited content model — e.g. "model content types for a B2B services site on Sitecore XM Cloud", "design a composable content model for a multi-brand retail DXP", or "turn this content audit into a clean type hierarchy".
-
Authors a comprehensive editorial governance document covering content lifecycle, roles and responsibilities, workflow stages (draft → review → approve → publish → archive), quality gates, compliance requirements, and taxonomy maintenance policies. Designed for DXP and headless CMS projects where multiple editorial teams or stakeholders share one publishing platform. Use when onboarding a new editorial team to a CMS, establishing governance before a platform migration go-live, or remediating inconsistent content practices — e.g. "write an editorial governance document for our Sitecore XM Cloud site", "define content workflow roles for our multi-brand Optimizely CMS", or "document publishing standards for our Umbraco site ahead of launch".
-
Defines a canonical metadata schema for a DXP or CMS project covering SEO, Open Graph, Twitter Card, structured data (JSON-LD Schema.org), localisation, personalization signals, and analytics data layer. Produces a field-by-field specification with required/optional flags, allowed values, character limits, and per-platform implementation notes (Sitecore, Optimizely, Umbraco, Contentful, Contentstack). Use when standardising metadata across a multi-template site, establishing SEO hygiene before migration go-live, or designing the structured data layer for a new DXP build — e.g. "build a metadata schema for our Sitecore XM Cloud site", "define SEO and JSON-LD fields for our Optimizely product pages", or "standardise metadata across 12 content types".
-
Builds a controlled-vocabulary taxonomy for a DXP or CMS project: hierarchical, flat, or faceted structures. Produces term lists, parent–child relationships, scope notes, usage guidelines, and a governance policy for term creation, maintenance, and retirement. Includes per-platform implementation notes (Sitecore, Optimizely, Umbraco, Contentful, Contentstack). Use when establishing the tagging and categorisation foundation for a new or re-platformed digital experience — e.g. "design a product taxonomy for our B2B site", "build a topic taxonomy for our editorial CMS", or "design faceted taxonomy for a resource library".
-
Generate on-brand page copy for any section of a digital experience — hero, value propositions, feature overviews, about, FAQs, calls to action — given a content brief and brand voice guide. Produces structured copy blocks with character counts, heading hierarchy, and CMS field mapping ready for Sitecore, Optimizely, Umbraco, or any headless front-end.
-
Author production-ready component microcopy — CTAs, form labels, placeholder text, inline help, error messages, empty states, success/warning/info toasts, confirmation dialogs, and onboarding tooltips — calibrated to the brand voice and the component's character-limit constraints. Produces a microcopy sheet ready for Figma annotation, design-system documentation, or CMS Dictionary items (Sitecore, Optimizely, Umbraco).
-
Write SEO-optimised page titles, meta descriptions, Open Graph tags, Twitter Card metadata, and canonical URLs for any page type — matching search intent, respecting character limits, and adapting to the CMS field structure of Sitecore, Optimizely, or Umbraco. Produces a metadata sheet ready for bulk CMS import or developer handoff.
-
QA a translated locale file or CMS export against the English source: detect untranslated strings, tone/register drift, placeholder corruption, character overflow against CMS field limits, terminology inconsistencies, and RTL/bidi issues. Produces a structured QA report with severity ratings and fix guidance for translators or localization engineers.
-
Transform English source copy into a locale-ready translation pack — JSON, XLIFF 1.2/2.0, or CMS-native format — for a target locale. Handles cultural adaptation notes, placeholder integrity, character expansion budget warnings, and CMS-specific export structure for Sitecore Language Fallback, Optimizely Language Branches, and Umbraco Language Variants / Dictionary items.
-
Scaffold a new brand theme on top of an existing base token set for house-of-brands and white-label DX projects. Produces a DTCG-compliant semantic override layer, a resolved token diff showing exactly what changes from the base, a theme-switcher integration snippet, and platform-specific injection instructions for Sitecore SXA, Optimizely, Umbraco, and headless front-ends. Keeps the primitive and component tiers untouched so maintenance stays centralised.
-
Generate production-ready component documentation covering anatomy, props/API table, interactive states, variants, design token callouts, accessibility requirements (WCAG 2.2 AA, ARIA patterns, keyboard interactions), and platform-specific usage notes for Sitecore SXA, Optimizely, Umbraco, and headless front-ends. Output is a structured Markdown file ready for Storybook, Supernova, zeroheight, or a Starlight-based design system site.
-
Produce design system governance documents: usage guidelines for a system section or specific component, contribution guide (how to propose changes, PR checklist, review process), version and deprecation policy, do/don't pattern library, and a design decision log (ADR-style). Works for both client-facing design system sites and internal RBA team governance. Adapts output for Starlight, Supernova, zeroheight, and plain Markdown wikis.
-
Audit an existing design token set for W3C DTCG spec compliance, three-tier architecture completeness (primitive → semantic → component), naming convention violations, missing token types, unresolved aliases, and cross-platform export readiness. Produces a severity-ranked finding report and a prioritised remediation plan. Adapts output for Sitecore SXA, Optimizely CMS, Umbraco, Tailwind, and Style Dictionary v4 pipelines.
-
Generate a complete, production-ready three-tier DTCG-compliant design token architecture (primitive → semantic → component) from a brand brief, Figma variables export, or design system specification. Outputs structured JSON files ready for a Style Dictionary v4 pipeline. Supports multi-brand house-of-brands setups with base + brand-override file separation. Adapts component coverage for Sitecore SXA, Optimizely, Umbraco, and headless front-ends.
-
Configure a Style Dictionary v4 build pipeline to transform DTCG-compliant token JSON into CSS custom properties, SCSS variables, JavaScript/TypeScript modules, Tailwind config, iOS Swift enums, and Android Kotlin objects — from a single token source. Supports multi-brand builds where each brand produces its own output set. Generates the full config file, build script, npm scripts, and README. Platform adapters for Sitecore SXA, Optimizely, Umbraco, and headless front-ends.
-
Static WCAG 2.2 AA accessibility gate for generated or hand-authored UI components. Validates contrast ratios against the project token layer, checks ARIA roles and states, verifies keyboard interaction model, inspects focus styles, and flags target-size violations. Produces a pass/fail verdict with a fix-per-finding list that a developer can resolve in the same coding session. Use before merge or QA handoff to prevent accessibility debt from entering the codebase.
-
Audit drift between Figma Variables (via the Figma MCP server) and the codebase token file. Reports tokens added, removed, renamed, or changed in value; flags contrast regressions introduced by Figma-side changes; and generates an updated token file and diff that a developer can review and merge. Use at the start of a sprint or after a design-system update to keep code and Figma in lockstep.
-
Export Figma Variables from the active file via the Figma MCP server to a W3C DTCG-compliant .tokens.json file, then transform it into CSS custom properties, Style Dictionary config, and optional framework-specific token outputs (Tailwind theme extension, SCSS variable map, .NET token class). Validates WCAG 2.2 AA contrast for all semantic color pairs at extraction time so the token layer ships accessibility-guaranteed before any component uses it.
-
Read a Figma frame or component via the Figma MCP server, extract design tokens, layout structure, and variant specifications, then generate an accessible, design-system-compliant component. Supports React/TypeScript, Razor/.NET, Astro, and Vue with Sitecore, Optimizely, and Umbraco rendering-layer adapters. All generated components include ARIA roles, keyboard interaction, and a contrast pre-check against the project token layer.
-
Synthesizes raw discovery material — interview transcripts, workshop notes, survey responses, or observation logs — into a structured insights report ready for client review. Apply when the user pastes or references interview transcripts, workshop notes, survey data, or stakeholder session outputs and asks to synthesize, analyze, theme, or report on them. Also applies when the user says "turn these notes into insights", "what themes came up across interviews", "write up the discovery", or "pull findings from these sessions". Works across CMS/DXP platform-selection, experience-design, and digital-strategy engagements. Skip when the user wants only a stakeholder map (use stakeholder-map), a root-cause chain (use five-whys-root-cause), or a persona (use persona-draft).
-
Drafts research-grounded personas from synthesized discovery data using a Jobs-to-be-Done backbone. Apply when the user has discovery data (interview synthesis, themes, quotes) and wants to produce audience personas, user archetypes, or content author profiles. Also applies when the user says "write a persona for", "create audience segments", "who are our users", or "draft an author persona". Produces JTBD-first personas, not demographic placeholders. Skip when the user has no discovery data and wants entirely hypothetical personas — ask for research input first.
-
Converts discovery insights, interview themes, and workshop outputs into a structured requirements register with MoSCoW prioritization and acceptance-criteria stubs. Apply when the user has discovery findings and wants to produce requirements, user stories, acceptance criteria, or a prioritized feature list for a DXP or CMS implementation. Also applies to "help me turn these findings into requirements", "write user stories from our discovery", "prioritize the requirements", or "build a requirements register". Skip when the user is still in the synthesis phase (use discovery-interview-synthesis first) or when requirements already exist and need only refinement (edit them directly).
-
Builds a structured stakeholder register and engagement plan from names, roles, or interview notes provided by the user. Apply when the user asks to map stakeholders, build a RACI, plan stakeholder engagement, identify who needs to be involved, or assess influence and interest across a project. Works for CMS/DXP implementations, platform-selection engagements, and digital-strategy workshops. Skip when the user wants raw interview synthesis (use discovery-interview-synthesis) or root-cause analysis (use five-whys-root-cause).
-
Facilitates a structured Five Whys root-cause analysis from a stated problem. Apply when the user describes a problem, pain point, or failure mode and wants to find the root cause — especially in digital-experience, content-operations, or platform-performance contexts. Also applies when the user asks to run a root-cause analysis, build a fishbone diagram, understand why something keeps happening, or trace a symptom back to its source. Skip when the user wants a synthesis of interview themes (use discovery-interview-synthesis) or stakeholder mapping (use stakeholder-map).
-
Authors an immutable Architecture Decision Record (ADR) following the MADR (Markdown Architectural Decision Records) format. Captures context, the decision drivers, all considered options with pros/cons, the decision, and its consequences. Handles superseding an existing ADR when the team changes course. Use whenever the team makes a significant technical decision — platform selection, integration pattern, auth scheme, infrastructure approach, coding standard, or any choice with lasting architectural consequences.
-
Generates a suite of architecture diagrams — C4 Context, C4 Container, sequence diagrams, data-flow diagrams, and MACH-layer views — using Mermaid syntax. Reads the repo or a described stack to produce accurate, audience-calibrated diagrams for client presentations, ADRs, solution design documents, or governance reviews. Works with any composable DXP stack with named adapters for Sitecore XM Cloud, Optimizely, Umbraco, Azure, and MACH composable.
-
Generates a developer or content-editor onboarding guide for an existing DXP project by reading the repo structure, stack configuration, CI/CD pipelines, and access patterns. Produces a structured Markdown guide covering environment setup, architecture overview, contribution workflow, key contacts, and common tasks — so a new joiner or client team can be productive on day one. Use at project kick-off, before a new team member joins, or as part of a client handover package.
-
Produces a complete operational runbook for any platform service or environment — including topology diagrams, access reference, standard operating procedures, incident playbooks with KQL queries, rollback procedures, and maintenance schedules. Reads repo IaC files and CI/CD pipelines to populate resource names and deployment steps accurately. Optionally queries the Azure CLI for live resource inventory. Use before go-live, during a managed-services onboarding, or when updating runbooks after infrastructure changes.
-
Generates a complete Solution Design Document (SDD) for a DXP feature, integration, or platform component by reading the repo, relevant tickets, and any existing architecture notes — rather than relying on the author's memory. Produces a structured, client-ready Markdown deliverable covering scope, architecture, data flows, component design, security, performance considerations, and open decisions. Use when beginning a significant feature, preparing for a design review, or documenting a completed sprint's technical decisions.
-
Audits existing documentation for staleness relative to current code, IaC, and recent git history, then updates or flags sections that are out of date. Detects documentation drift across READMEs, architecture docs, SDDs, runbooks, ADRs, and onboarding guides. Produces a staleness report and, where changes are unambiguous, applies in-place updates. Use at sprint end, before a release, or during a client handover to ensure docs reflect the current state of the platform.
-
Generates a slide-by-slide script for the RBA Envision recommendations deck — the phase's most important deliverable — where every slide title states a FINDING or CONCLUSION, never a topic. Follows the Context → Insights → Concepts → Recommendation → Next-Steps arc on the RBA master PowerPoint template, one insight per slide, with a three-reasons-why recommendation. Apply when the user has Envision research synthesized and asks to build the Envision deck, findings deck, recommendations deck, or readout, or says "present the research" or "build the leadership deck". Skip when the user wants a kickoff (use project-kickoff-deck) or interview questions (use stakeholder-interview-guide).
-
Generates a complete slide-by-slide script for a 60-90 minute RBA project kickoff, ordered against the RBA master PowerPoint template's layouts and boilerplate slides (project cadence, risks, dependencies, how-we-work-together, PM deliverables). Apply when the user is starting an engagement and asks to build a kickoff deck, kickoff presentation, or kickoff slides, or says "prep my kickoff" or "build the kickoff for [client]". Produces a 12-14 slide script with owner-and-date next steps that the user pastes into Claude in PowerPoint to assemble inside the master template. Skip when the user wants an interview guide (use stakeholder-interview-guide) or a research findings deck (use envision-findings-deck).
-
Builds a tailored RBA Envision stakeholder interview guide for one stakeholder type — warm-up, themed core questions with follow-up prompts, standard closing questions, and a post-interview synthesis-notes section — and exports it to a branded Word doc. Apply when the user is preparing Envision-phase discovery interviews and asks for an interview guide, a discussion guide, stakeholder questions, or says "write interview questions for [role]", "build a discovery guide", or "prep for my stakeholder interview". One guide per stakeholder TYPE, not per person. Skip when the user wants to synthesize completed interviews (that is downstream) or build a kickoff/findings deck (use project-kickoff-deck or envision-findings-deck).
-
Produces a structured project estimate (hours by role and phase) from a brief or requirements summary; use when a delivery lead, PM, or BA needs to size a new engagement or proposal before a SOW is drafted.
-
Drafts a formal Change Order when project scope changes mid-engagement, capturing delta scope, delta effort, revised timeline, and fee adjustment, referencing the original SOW; use when a client requests additional work, a scoped item expands beyond the SOW boundary, or an assumption in the SOW has proven incorrect.
-
Drafts a complete, client-ready Statement of Work from an approved estimate and WBS, populating all standard SOW sections including scope, deliverables, assumptions, exclusions, timeline, pricing, RACI, acceptance criteria, and change-order clause; use when a proposal has been approved and a formal contract document is needed.
-
Generates a hierarchical Work Breakdown Structure (WBS) from an approved estimate or project brief, decomposed into phases, workstreams, tasks, and sub-tasks with owner roles and milestone gates; use before kicking off a project or populating Azure DevOps / a project plan.
-
Applies structured risk weighting to a raw project estimate, scoring each line item across complexity, integration, third-party dependency, and stakeholder-readiness axes, then adjusts hours and surfaces a risk register; use after build-estimate and before draft-sow to validate the estimate holds up under scrutiny.
-
Compares a post-incident corrective-action list or postmortem against an existing environment runbook and produces a targeted diff — adding missing playbook steps, updating stale KQL queries, closing escalation-path gaps, and versioning the runbook. Designed to run immediately after every postmortem to keep living runbooks current.
-
Drafts a concise, structured incident summary from raw logs, alert history, engineer notes, or a triage card. Produces two artefacts: an internal technical summary (for the war room and ITSM ticket) and a client-facing status update (plain language, no internal jargon). Covers the full detection-to-resolution timeline.
-
Analyses Azure Monitor alert payloads, Application Insights exceptions, or raw log excerpts to produce a structured triage card — severity classification, blast-radius estimate, likely root cause hypotheses, and a first-response checklist. Works across Sitecore, Optimizely, Umbraco, and headless DX stacks on Azure.
-
Facilitates a blameless post-incident review for a resolved DX platform incident. Produces a complete postmortem document: timeline, five-whys root cause analysis, contributing factors, corrective actions with owners and due dates, and a lessons-learned section. Follows Google SRE and ITIL-aligned blameless principles.
-
Audits an Azure Monitor alert ruleset for a DX environment — identifying coverage gaps, alert fatigue sources (over-sensitive thresholds, duplicate rules, missing deduplication), broken action groups, and KQL query errors. Produces a prioritised improvement plan with ready-to-apply Bicep or CLI commands for each fix.
-
Produces a complete OpenAPI 3.1 or AsyncAPI 3.0 specification for a new DXP integration endpoint — including authentication scheme, rate-limit headers, error schema, pagination contract, and example payloads. Use when starting a new integration between any two systems (CMS, CRM, commerce, CDP, search) and needing an agreed-upon contract before a single line of implementation code is written.
-
Generates a typed, production-ready API client (TypeScript or C#) for an integration endpoint, wired to Azure API Management with retry decorators, bearer-token refresh via MSAL, structured logging, and OpenTelemetry trace propagation. Use after integration-api-spec has produced the contract and you need the implementation scaffold — not the logic, just the reliable plumbing every integration client needs.
-
Authors a canonical data-mapping contract (source schema → canonical model → target schema) with field-level transformation rules, null/missing handling, type coercions, and a machine-readable JSON Schema or TypeScript type that both integration sides can import. Use when two systems need to exchange data and the field shapes, naming conventions, or data types differ — which is always.
-
Designs the complete error-handling and resilience layer for a DXP integration: HTTP/event status taxonomy, retry and back-off policies, circuit-breaker thresholds, dead-letter queue routing, idempotency key strategy, and poison-message handling. Use after the API spec and client scaffold exist — this skill turns the "what happens when it breaks" question into concrete, testable code patterns.
-
Defines the observability baseline for a DXP integration: Azure Application Insights KQL queries, distributed trace propagation configuration, alert thresholds, a Kusto dashboard definition, and an on-call runbook template. Use after the client scaffold and resilience design are in place — observability is the last piece before a new integration goes to staging.
-
Packages a tested RAG knowledge base assistant for deployment: generates production-ready Bicep infrastructure as code, an Azure Bot Service + Teams manifest for channel publishing, an OpenAPI tool definition for Copilot Studio or Claude plugin integration, a client-handover summary, and an operator quickstart. Use after the assistant has passed the kb-qa-evaluator gate — "package our HR policy assistant for Teams", "generate the Bicep IaC for our client KB deployment", or "produce a Copilot Studio plugin definition for our product knowledge assistant".
-
Runs a structured evaluation harness against a deployed or local RAG knowledge base assistant: generates a golden question set, scores answers for faithfulness (groundedness), answer relevance, context precision, and hallucination presence, then produces a pass/fail gate report with actionable remediation. Use before promoting a RAG pipeline to production, after index re-ingestion, or as part of a scheduled quality review — "evaluate our KB assistant before launch", "run a faithfulness check after we updated the SharePoint index", or "generate a golden Q&A set for our policy assistant".
-
Designs the full retrieval-augmented generation (RAG) pipeline for a knowledge base assistant: retrieval strategy (vector, hybrid, agentic), re-ranking, prompt template, answer grounding rules, and security architecture. Produces an Azure AI Search + Azure OpenAI reference architecture with Bicep infrastructure stubs and a system prompt template ready for evaluation. Use after the ingestion spec is complete — "design the RAG pipeline for our ingested SharePoint KB", "what retrieval strategy suits our multi-source assistant?", or "write the system prompt for our internal policy assistant".
-
Audits raw content sources for a knowledge base assistant — SharePoint sites, Sitecore/Optimizely/Umbraco CMS exports, Azure Blob Storage document libraries, or file system dumps — and produces a complete ingestion pipeline specification: source connector config, chunking strategy, metadata enrichment schema, and index field mapping for Azure AI Search. Use when starting a RAG assistant build: "plan the ingestion pipeline for our SharePoint intranet", "how do I chunk our Sitecore product docs?", or "design the indexer for this PDF library".
-
Produces a day-2 operations runbook for a live knowledge base assistant: index re-ingestion cadence and procedures, content drift monitoring, cost and token usage dashboards, performance SLA tracking, incident escalation paths, and rollback procedures. Use when handing over a deployed assistant to a client ops team, setting up managed services support, or formalising maintenance responsibilities — "write the runbook for our client's SharePoint KB assistant", "create the ops guide for our deployed RAG pipeline", or "document the re-ingestion procedure".
-
Produces a structured competitor teardown brief covering technology stack detection, heuristic/UX review (Nielsen 10), SEO snapshot, content strategy read, and a digital maturity score. Apply when a user provides a competitor domain or name and asks for a competitor analysis, teardown, audit, or platform read. Works for any sector; surfaces platform-specific implications when Sitecore, Optimizely, Umbraco, or composable DXP signals appear. Skip when the user wants only a Five Forces analysis (use five-forces-analysis), only an SEO gap (use seo-content-gap), or only a maturity benchmark (use digital-maturity-benchmark).
-
Scores a client's (or competitor's) digital maturity across six dimensions — experience design, content operations, data/personalization, commerce capability, cloud/infrastructure, and accessibility — against an RBA baseline and inferred sector quartiles. Produces a scored benchmark report with evidence citations and a prioritized improvement roadmap. Apply when a user asks for a digital maturity assessment, maturity scoring, benchmark, or "where does the client sit vs. sector peers?". Works as a standalone engagement or as a companion to competitor-teardown. Skip when the user wants a competitor teardown (use competitor-teardown) or a full SEO/content gap analysis (use seo-content-gap).
-
Applies Porter's Five Forces adapted for digital markets — competitive rivalry, threat of new entrants, supplier power, buyer power, and threat of substitutes — with a DXP/CMS lens for platform-selection and market-entry contexts. Apply when a user asks for a Five Forces analysis, industry/market structure assessment, or competitive landscape framing. Produces a scored, cited brief with strategic implications for RBA's client. Skip when the user wants a single-competitor teardown (use competitor-teardown) or a market sizing exercise (use market-sizing).
-
Produces a cited TAM/SAM/SOM market sizing model using both top-down (analyst reports, industry data) and bottom-up (addressable account estimation) approaches. Compares convergence between methods and delivers a model with labeled assumptions. Apply when a user asks to size a market, calculate TAM/SAM/SOM, or estimate addressable opportunity for a client, vertical, or platform. Skip when the user wants a Five Forces structural analysis (use five-forces-analysis) or a competitor teardown (use competitor-teardown).
-
Produces a prioritized SEO and content gap register by comparing a client domain against 2–5 competitor domains: keyword overlap, topic authority gaps, backlink delta, Core Web Vitals comparison, and AI-search visibility (ChatGPT/Perplexity citation signals). Apply when a user asks for a content gap analysis, keyword gap, SEO competitor comparison, or "what topics do we not own that competitors do?". Skip when the user wants a full competitor teardown (use competitor-teardown) or a market sizing exercise (use market-sizing).
-
Extracts every action item from a normalized meeting record or raw notes block, assigns owners and due dates, validates against named participants, and outputs a complete action register in markdown table and optional CSV format. Apply when the user asks to "pull the action items," "who owns what," "build the action register," or "track the follow-ups from this meeting." Also applies when reviewing whether prior meeting actions are complete before writing a follow-up. Skip when the user wants to push actions to Jira directly — chain to jira-confluence-push after this skill.
-
Extracts decisions from a normalized meeting record (from meeting-transcript-parse) or a raw notes block and formats them as structured decision log entries using a DACI model (Driver, Approver, Contributor, Informed). Outputs entries ready to append to a Confluence project decision log or a standalone markdown register. Apply when the user asks to "build the decision log," "log the decisions from this meeting," "add to the decision register," or "document what was decided." Also applies when a steering committee or governance meeting produced explicit decisions that need to be permanently recorded. Skip when no decisions were made — use action-item-tracker for meetings that are purely task-oriented.
-
Pushes structured action items and decision log entries to Jira (create/update issues) and Confluence (append to a meeting notes page or decision register) via the Atlassian MCP server or REST API. Also supports Azure DevOps Boards as an alternative work-item backend. Apply when the user asks to "push to Jira," "create the tickets," "update Confluence," "log the decisions in Confluence," or "sync the action register." Requires ATLASSIAN_API_TOKEN and ATLASSIAN_BASE_URL environment variables (or AZURE_DEVOPS_PAT for ADO). Skip when the user only wants a local markdown output — use action-item-tracker or decision-log-builder for that.
-
Drafts a polished stakeholder follow-up email or Microsoft Teams message from a normalized meeting record and/or action register. Restates decisions, lists assigned actions with owners and deadlines, flags unresolved questions, and matches RBA's professional but direct communication style. Apply when the user asks to "draft the follow-up," "write the meeting email," "send the summary," or "post the recap to Teams." Works for both internal and client-facing follow-ups. Skip when the user wants to push to Confluence instead of email — use jira-confluence-push for that.
-
Ingests a raw meeting transcript or rough notes — from Microsoft Teams, Zoom, a recording auto-caption export, or manual copy-paste — and produces a clean, normalized meeting record with participants, agenda items covered, discussion threads, decisions, action items, unresolved questions, and parking-lot items. Apply when the user pastes or references a transcript, notes file, or meeting recording export and asks to "clean up," "summarize," "pull actions from," or "write up" a meeting. Works for internal RBA team meetings, client status calls, steering committees, sprint reviews, and discovery workshops. Skip when the user already has a structured record and only wants to build a decision log (use decision-log-builder) or push to Jira (use jira-confluence-push).
-
Generates a full-depth project onboarding document from repo context — reads CLAUDE.md, installed plugin manifests, CI/CD config, IaC files, environment variable templates, and git history to produce a role-specific Markdown guide covering access, environment setup, architecture, contribution workflow, common tasks, and troubleshooting. Supports developer, BA, QA, content-editor, and client-handover audiences. Use at project kick-off, when a new team member joins, or to refresh a stale guide before client handover.
-
Generates a sprint-by-sprint "first 30 days" ramp plan for a new team member joining an in-flight RBA DXP project. Reads the current backlog, installed skills, active sprint, branching history, and project conventions to produce a personalised schedule covering shadow work, first solo tasks, and ownership milestones. Produces a Markdown ramp plan with weekly goals, specific tickets to shadow or own, and skill checkpoints. Use when a new hire, contractor, or transferred team member joins a project mid-flight.
-
Audits the installed Claude Code skills and plugin catalog against the project's actual delivery needs. Reads the current plugin manifests, CLAUDE.md, and recent git history to identify: gaps (skills that would accelerate delivery but aren't installed), redundancy (overlapping skills that could be rationalised), and mis-configuration (skills granted too-broad tool permissions). Produces a prioritised install recommendation with rationale and least-privilege tool guidance. Run before a new sprint cycle or when adding a new team member.
-
Generates a targeted self-study ramp-up path for a specific technology in the project stack (e.g. Sitecore XM Cloud headless, Umbraco 13 + Delivery API, Azure Container Apps, Coveo Atomic). Produces a curated, sequenced learning plan — official docs, RBA-relevant patterns, hands-on exercises, and validation checkpoints — tuned to what is actually in use on the project. Not a generic reading list; each step builds on the last and connects directly to the project codebase. Use when an engineer picks up an unfamiliar technology mid-engagement.
-
Comprehensive image, video, font, and static-asset review for RBA DX projects. Checks format selection (WebP/AVIF), compression, responsive srcset, lazy loading, CLS-safe sizing, and delivery pipeline configuration for Sitecore Media Library, Umbraco ImageSharp, Optimizely DXP media, and Next.js Image. Produces findings with estimated LCP/CLS impact and code-level fixes for each asset type.
-
Convert any performance audit report (from perf-cwv-audit, perf-caching-review, perf-render-blocking-analysis, or perf-asset-optimisation) into a prioritised, sprint-ready backlog. Groups findings by impact and component, estimates effort, assigns owner (engineering/design/devops/content), and produces Azure DevOps work items or Jira stories with acceptance criteria tied to specific CWV thresholds. Includes a quick-win list and a governance epic for CI budget enforcement setup.
-
Define, encode, and enforce web performance budgets in CI. Generates a Lighthouse CI configuration (lighthouserc.json) with CWV assertion thresholds, a GitHub Actions workflow that runs LHCI on every pull request, and performance budget config for bundle-size checks (bundlesize / size-limit). Produces a PR comment with metric comparisons against the baseline. Targets Azure DevOps pipelines as an alternative output format.
-
HTTP cache-header audit across CDN, CMS output cache, .NET API layer, and browser cache. Inspects Cache-Control, ETag, Vary, and CDN-specific headers on origin responses. Maps each resource type to the correct caching strategy and identifies gaps that force unnecessary origin hits. Covers Azure CDN, Azure Front Door, Sitecore output cache, Optimizely DXP CDN, and Umbraco output/Redis caching.
-
Full Lighthouse + CrUX audit for Core Web Vitals (LCP, INP, CLS). Combines automated Lighthouse scores with field data from Chrome User Experience Report, diagnoses root causes per metric, and produces a severity-ranked finding report with estimated impact deltas. Adapts analysis for Sitecore XM Cloud, Optimizely Content Cloud, Umbraco, and headless/composable stacks on Azure.
-
Deep-dive analysis of render-blocking CSS, JavaScript, and font resources that delay First Contentful Paint and Largest Contentful Paint. Examines <head> resource order, async/defer discipline, critical CSS coverage, font preloading, and third-party script loading strategies. Produces prioritised fixes with estimated FCP/LCP savings. Platform-aware for Sitecore, Optimizely, Umbraco, and Next.js.
-
Defines behavioral, firmographic, lifecycle-stage, and real-time session audiences for personalization and experimentation targeting. Produces platform-ready segment definitions for Optimizely (ODP/Audience Builder), Sitecore CDP (batch and real-time), and Adobe Target — with adapter notes for Segment, mParticle, and custom CDPs.
-
Assesses an organization's personalization and experimentation program against a five-stage maturity model — covering hypothesis quality, experiment velocity, statistical rigor, organizational alignment, and platform utilization — then outputs a scored assessment and a prioritized improvement roadmap.
-
Transforms quantitative signals (analytics drop-offs, heatmaps, session recordings) and qualitative signals (user research, surveys, stakeholder input) into a prioritized backlog of evidence-backed, statistically testable personalization and A/B test hypotheses. Uses the Problem–Solution–Result (PSR) framework enforced by RBA's experimentation standards.
-
Authors a complete measurement specification for a personalization or experimentation initiative — north-star metric, primary and secondary success metrics, guardrail metrics, counter-metrics, and data-collection requirements — so results are unambiguous before the experiment or personalization rule launches.
-
Designs a complete A/B, multivariate, or feature-flag experiment from an approved hypothesis — covering traffic split, sample size calculation, run duration, variant descriptions, quality assurance checklist, and platform-specific setup instructions for Optimizely, Sitecore CDP/Personalize, and Adobe Target.
-
Reads an existing RBA Skills plugin bundle (or standalone skill) and produces a gap report listing every violation of the CLAUDE.md rules: slug/name identity mismatches, missing required files, invalid facet values, bare-Bash or wildcard-WebFetch privilege violations, SKILL.md line-count overruns, and custom RBA fields leaking into plugin.json. Use before opening a PR to catch issues before CI does.
-
Generates a subagent definition (.md file) for an existing RBA Skills plugin bundle, building a correct `tools` list under least-privilege rules, writing an orchestration step sequence, and confirming the agent slug matches its filename. Use when you need an autonomous reviewer or orchestrator agent to accompany a plugin's skills.
-
Generates hooks/hooks.json for an existing RBA Skills plugin, classifying each hook event by security risk level, explaining what each hook will run and when, and refusing to write hooks that access credentials or execute with elevated privilege. Use when you need to automate post-write formatting, pre-commit validation, or stop-event notifications within a plugin's lifecycle.
-
Scaffolds a complete, CI-passing RBA Skills plugin bundle from a plain-language description. Generates all four required files (plugin.json, rba-meta.json, README.md, first SKILL.md) plus optional agents, commands, and hooks. Enforces slug=name identity, least-privilege allowed-tools, facet assignment, and the 500-line SKILL.md limit. Use when starting a new plugin from scratch.
-
Adds a single, CI-passing skill (SKILL.md plus optional reference.md, templates.md, examples.md) to an existing RBA Skills plugin bundle or standalone path. Derives the correct slug, enforces least-privilege allowed-tools, caps SKILL.md at 500 lines, and routes depth to sibling files. Use when you want to extend an existing plugin with a new skill or create a standalone skill without the full plugin scaffold.
-
Sets up an RBA budget tracker — budget dashboard, weekly hour-burn log, and by-role summary with traffic-light status — pre-populated and ready for the team to update; use at kickoff and refresh weekly.
-
Writes an RBA project closure report — objectives vs. outcomes, budget and schedule variance, lessons learned, open items, and the AE handoff; use within one week of final delivery.
-
Stands up an RBA RAID log — a 5-tab Risks / Issues-Actions / Decisions / Lessons-Learned / Contract-SOW workbook plus a reference tab — pre-populated with standard project risks; use at kickoff and review weekly.
-
Builds a focused RBA RASCI chart — 8 to 12 key activities, exactly one Accountable per activity — that makes ownership explicit from day one; use at kickoff to align RBA and client roles.
-
Builds an RBA project schedule workbook — milestone tracker, Gantt-style phase timeline, and team allocation — from SOW phases, team, and key dates; use at kickoff to baseline a schedule and weekly to keep it current.
-
Builds a Requirements Traceability Matrix (RTM) that maps each acceptance criterion in a story backlog to test IDs in your Gherkin or Playwright test suite, surfaces coverage gaps, and generates a release-gate checklist; use before a sprint review, UAT hand-off, or go-live to prove no AC is untested.
-
Performs systematic boundary-value, negative-path, concurrency, and accessibility edge-case analysis on a user story or feature area, producing a prioritised inventory that qa-gherkin-scenarios can use to close coverage gaps; use when stories are accepted but before test writing to avoid happy-path bias.
-
Parses a user story, acceptance criteria block, or PRD section into a structured Gherkin feature file with Given-When-Then scenarios, tagged by persona, risk, and platform; use when converting requirements into executable BDD spec before or during a sprint.
-
Generates typed Playwright TypeScript test scripts from a Gherkin feature file or a list of scenarios, using resilient selectors (role → label → testId → CSS), Page Object Model classes, and fixture patterns; use after qa-gherkin-scenarios or when an existing feature file needs automation.
-
Produce a polished, plain-language client-facing feature announcement from the harvest change set — benefit-led, jargon-free, suitable for a Confluence page, email update, or slide deck. Strips internal refactors, infrastructure changes, and commit SHAs.
-
Generate a semver-aligned, conventional-commit categorised CHANGELOG.md for engineers — with breaking changes prominently flagged, security fixes called out, and migration notes auto-drafted. Reads the harvest.json produced by release-notes-harvest.
-
Harvest and categorise commits, pull requests, and work items from Git or Azure DevOps into a structured change set — the prerequisite step before generating any audience-specific release note. Outputs a machine-readable JSON manifest and a human-readable summary.
-
Write a project-manager sprint or release summary — delivery metrics, decision log, risk callouts, open items, and next-sprint commitments — synthesised from the harvest change set and Azure DevOps sprint data. Aimed at PMs, delivery leads, and internal stakeholders.
-
Publish a ready-to-ship release note document to Confluence, Slack, or generate a formatted email draft — reads from .release-notes/ drafts produced by the rendering skills and posts to the specified target with channel-appropriate formatting.
-
Synthesises the draft RFP response into a punchy, win-theme-focused executive summary calibrated to the evaluator's stated priorities and any page or word limit; use after all response sections are drafted.
-
Parses an inbound RFP document and extracts every stated, implied, and evaluation requirement into a structured requirements register; use at the start of every bid to give the whole team a single, authoritative source of what needs to be answered.
-
Drafts a specific RFP response section by pulling from the local answer library and tailoring it to the client context and evaluation criteria; use for each section in the compliance matrix after owners have been assigned.
-
Runs a go/no-go qualification scorecard against an inbound RFP opportunity — win probability, strategic fit, resource cost to bid, and a clear recommendation; use within 24 hours of receiving an RFP before committing to a full response.
-
Matches an RFP's requirements and evaluation criteria against RBA's case study library and returns the top 3–5 most relevant proof points with a recommended narrative angle for each; use after rfp-shred to inform experience sections.
-
Generates a full compliance matrix from an RFP requirements register — maps every requirement to a response location, assigns an owner, records compliance status, and flags risks; use immediately after rfp-shred at bid kick-off.
-
Writes a complete slide-by-slide script for the RBA master PowerPoint template — duplicating the pre-built About RBA block, applying engagement-model rules, and using point-first slide titles. Output is pasted into Claude-in-PowerPoint to build a branded capability or proposal deck (PPTX).
-
Drafts a 1–2 page tailored executive briefing (DOCX-ready) that connects a trend to a specific client's situation and ends with a conversation starter, not a pitch. RBA's between-projects relationship tool — use to keep a warm account warm or open a strategy conversation.
-
Produces a single-page RBA leave-behind (PDF-ready): navy header band, max three body sections, two- or single-column layout. Enforces the hard one-page rule — cut content, not margins. Use for event handouts, email attachments, and capability leave-behinds.
-
Drafts a concise RBA RFI response (DOCX-ready, under four pages): copies the prospect's questions verbatim, answers each directly, backs every claim with a specific engagement, and ends with one clear next step. For early-stage vendor evaluation — not a formal scored RFP.
-
Scans NuGet (packages.lock.json / *.csproj) and npm (package-lock.json / yarn.lock) dependency trees for known CVEs using dotnet CLI and npm audit. Grades findings by CVSS severity, distinguishes direct from transitive vulnerabilities, and produces an upgrade path for every affected package. Use weekly in CI, before go-live, and whenever a new package is added.
-
Audits consent mechanisms, cookie banners, PII handling, data-subject rights flows, and privacy policy completeness against GDPR and CCPA requirements. Covers front-end consent UX, server-side PII data flows, third-party tag/pixel governance, and Sitecore/Optimizely/Umbraco CMS-specific data handling. Produces a gap analysis and a prioritised remediation checklist.
-
Produces a polished, non-technical executive security summary suitable for client sign-off, governance review, or project handover. Translates OWASP, CVE, headers, and privacy findings into business-risk language with a RAG-status dashboard, risk narrative, remediation timeline, and residual-risk statement. Tone: professional, confident, no unnecessary alarm.
-
Reviews HTTP response headers against the OWASP Secure Headers Project recommendations: CSP, HSTS, X-Content-Type-Options, Referrer-Policy, Permissions-Policy, and more. Reads configuration from web.config, appsettings.json, Next.js config, nginx.conf, or Azure Front Door rules. Outputs a graded headers report and production-ready config snippets to close every gap.
-
Synthesises findings from owasp-security-audit, dependency-cve-scan, security-headers-audit, and privacy-consent-review into a single prioritised, sprint-ready remediation backlog. Deduplicates overlapping findings, assigns RICE-style effort estimates, groups by sprint, and generates Azure DevOps work-item YAML ready to import. Use after running any combination of the other security skills, or to triage an existing security report.
-
OWASP Top 10:2025-aligned source-code security audit of a file, directory, or git diff. Produces a severity-ranked findings report with CWE references, exact vulnerable code snippets, and platform-specific remediation steps for Sitecore, Optimizely, Umbraco, and .NET/TypeScript stacks. Use as a pre-merge gate on any change touching auth, data access, APIs, or third-party integrations.
-
Analyze existing content against GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) best practices — TLDR-first structure, expert quotes, inline citations, entity clarity, FAQ extraction — and produce a rewrite brief or direct rewrite that improves citation probability in ChatGPT, Perplexity, Google AI Overviews, and Microsoft Copilot. Optionally generate an llms.txt file for the site.
-
Identify keyword and topic gaps between your site and competitors using crawl exports, keyword ranking CSVs, and GSC performance data — score each opportunity by search volume × keyword difficulty × business value — and produce a prioritised content brief backlog ready for sprint planning. Use at the start of a content strategy engagement or as a quarterly operated-site opportunity review.
-
QA a redirect map before and after a DXP platform migration or site redesign — validate URL mapping coverage, detect redirect chains and loops, verify correct redirect types (301 vs. 302), check for missing high-priority URLs, and output deployment-ready redirect rulesets for IIS, Azure Front Door, Nginx, and Vercel. Use during pre-launch QA and post-launch monitoring sprints.
-
Audit existing JSON-LD schema markup against Google Rich Results requirements and AI answer-engine citation standards (ChatGPT, Perplexity, Google AI Overviews) — then generate corrected or missing JSON-LD for Article, FAQPage, BreadcrumbList, Organization, Product, HowTo, LocalBusiness, and SiteLinksSearchBox. Use before launch, after CMS migration, or as part of a GEO/AEO readiness review.
-
Full technical SEO audit from a crawl export or live URL — metadata completeness, canonicalisation, robots.txt and sitemap health, Core Web Vitals (LCP, INP, CLS), crawl budget, hreflang, internal linking depth — outputting a prioritised P1/P2/P3 finding list with remediation guidance. Use at pre-launch, post-migration, or as a quarterly managed-services health check.
-
Assembles a complete escalation package for a P1/P2 incident — incident timeline, business impact statement, actions-taken log, current hypothesis, and a structured escalation brief in the format expected by RBA engineering leads, vendor support portals (Sitecore, Umbraco, Optimizely, Microsoft), or client executive contacts. Use when an incident is not resolving within SLA or requires specialist/vendor engagement.
-
Converts a resolved support ticket into a reusable knowledge-base article — symptom description, root cause, step-by-step resolution, prevention guidance, and retrieval tags. Outputs a formatted article ready to post to Confluence, Zendesk Guide, or JSM Knowledge Base. Use after closing any P1–P3 ticket to build institutional knowledge and reduce repeat tickets.
-
Investigates the technical root cause of a support ticket by correlating error messages, logs, platform configuration, and the KB. Returns a structured RCA with evidence chain, confidence level, and fix recommendations. Use when a ticket is stalled, symptomatic, or recurring — or as the investigation step before drafting a resolution response. Supports Sitecore XM Cloud, Umbraco, Optimizely, and Azure-hosted DX platforms.
-
Scans the open support ticket queue in Zendesk or JSM and produces an SLA health dashboard — breached, at-risk, and healthy ticket counts by priority; dwell-time analysis; repeat-issue flags; and specific recommended actions. Use at the start of each support day or during an on-call handover to get immediate situational awareness across the queue.
-
Drafts professional client-facing responses for support tickets at any stage — acknowledgement, diagnostic update request, resolution summary, or closure. Calibrates tone to SLA elapsed time and ticket priority. Integrates with Zendesk and JSM to read ticket history and post the response. Use after triaging a ticket or when an engineer needs to communicate with the client.
-
Reads one or more support tickets (from Zendesk, JSM, or pasted text) and produces a structured triage record — ITIL priority (P1–P4), category, affected platform, recommended SLA targets, first-responder action, and routing. Use at the start of any incoming ticket workflow. Integrates with Zendesk and JSM MCP servers when available.
-
Drafts a single experiment brief: hypothesis (we believe / will result in / measured by), primary and guardrail metrics, audience, MDE and sample-size sanity check, variants, and stop criteria, for Optimizely Web/Feature Experimentation or VWO. Use when defining one A/B or MVT test.
-
Generates WCAG-conformant alternative text for images, distinguishing informative, decorative, functional, and complex (charts/diagrams) cases and flagging text-in-image. Use when a designer, content author, or engineer needs concise accurate alt text for a specific image rather than a full accessibility audit. Invoke with an image path, URL, or description; returns ready-to-paste alt text plus a rationale and any CMS-specific guidance for Sitecore, Optimizely, Umbraco, or headless frontends.
-
Generates mock API responses and a runnable stub (MSW handler, JSON Server config, or Azure Static Web Apps mock) from an OpenAPI spec or sample payload, so front-end work can proceed before the backend is ready. Use when unblocking a single integration during parallel development.
-
Drafts professional client-facing follow-up emails in RBA brand voice for a specific situation: scope-change confirmation, blocked-on-client nudge, missed-deadline notice, payment/invoice reminder, or escalation. Use when a PM or account lead needs one tactful, on-brand client message.
-
Scaffolds a single front-end component — React/Next.js vanilla, Sitecore JSS rendering, Sitecore SXA rendering, or an Umbraco block — complete with typed props, a Storybook story, a Vitest/Jest test stub, and accessible markup aligned to the project's existing conventions. Use when an engineer needs one new component wired to its CMS data source, not a whole design system. Adapters for Sitecore JSS/SXA and Umbraco headless make this RBA-specific.
-
Models a single content type / template as fields with types, validation, references, and reuse notes, then emits it for the target CMS: Sitecore template, Umbraco Document Type, Optimizely content type, or platform-agnostic schema. Use when defining one content type during content modeling — not the whole information architecture. Trigger phrases: "model this content type", "define the fields for [type]", "emit a Sitecore template for [type]", "create a Document Type for [type]", "scaffold an Optimizely content type", "give me the schema for [type]", "what fields should [type] have", "add [field] to [type]".
-
Inventories cookies and trackers on a page and maps each to a consent category (strictly necessary, functional, analytics, marketing) with purpose, provider, duration, and GDPR/CCPA notes, ready for a CMP configuration. Use when building or auditing cookie disclosure for one site.
-
Scaffolds a new Agent Skill that is spec-correct against Anthropic's open Agent Skills baseline and Claude Code's layered extensions. Use when the user asks to create, scaffold, design, author, build, or make a new skill, or describes a repeated workflow they want to capture as reusable. Applies an exploration-first intake methodology when the brief is ambiguous—ask before answering, surface assumptions, track insights, resist solutioning (full method in references/intake-methodology.md)—then drafts SKILL.md plus confirmed references/, scripts/, assets/ subdirs in one of three scopes—project (.claude/skills/[name]/), personal (~/.claude/skills/[name]/), or portable (open-baseline frontmatter for the Claude API or claude.ai Projects). Runs a verification gate covering frontmatter, body length, supporting-file references, and copyright-footer placement before reporting done. Skip when modifying an existing skill in place, answering questions about how skills work, or invoking a different existing skill.
-
Translates a cron expression (or Azure Logic Apps / Functions timer NCRONTAB) to plain English and back, validates it, and shows the next several run times in a chosen timezone. Use when configuring or auditing a single scheduled job.
-
Runs a fast Core Web Vitals triage on a single URL using PageSpeed Insights / CrUX data, ranks the top LCP, CLS, and INP offenders, and returns prioritized quick wins. Use when someone needs a rapid CWV read on one page, not a full performance audit.
-
Scans a codebase for environment variables and config keys, then produces a documented .env.example and a settings table (name, purpose, required, default, secret?, where set) for App Service / Key Vault. Use when documenting or onboarding the configuration surface of one project.
-
Diagnoses a stack trace or error message (Sitecore, Umbraco, Optimizely, .NET, Node, Azure), explaining the likely cause, where to look, and concrete fixes ranked by probability. Use when an engineer or support agent is stuck on one specific error.
-
Plans a single feature flag: naming, default state, targeting rules, kill-switch, cleanup/expiry plan, and rollout stages, emitted for Azure App Configuration, LaunchDarkly, or Optimizely Feature Experimentation. Use when introducing or auditing one flag.
-
Specs a single web form — fields, types, validation rules, required/optional, error and success copy, accessibility (labels, aria, focus order), spam protection, and submission destination (Dynamics 365, HubSpot, Power Automate, Sitecore Forms, Umbraco Forms). Use when defining one form before build: "spec the contact form", "write the form spec for [FormName]", "what fields does the [FormName] need", "define validation for [FormName]", "how should the [FormName] submit".
-
Writes a Conventional Commits message or a complete pull request description (summary, change list, test notes, risk, rollback) from the current diff. Use when an engineer needs a clean commit or PR write-up, not a code review.
-
Converts an RBA Harvey Ball vendor-comparison Excel scorecard into a fully branded, multi-slide PowerPoint deck — executive podium summary, capability matrix with Harvey-ball overlays, TCO pricing, and auto-grouped question-level detail pages. Use when someone uploads a Harvey Ball .xlsx scorecard and asks to make a deck, PPTX, presentation, or slides from it. Pairs with platform-selection / vendor-evaluation scorecards (Sitecore vs. Optimizely vs. Umbraco, SaaS vs. PaaS, etc.).
-
Plans responsive image delivery for a given asset or component: formats (AVIF/WebP fallback), srcset/sizes breakpoints, lazy-loading strategy, CDN transform parameters (Azure CDN, Cloudflare, Sitecore Media, Umbraco ImageSharp), and dimension/weight budgets. Use when optimising imagery for one component or page — distinct from performance-auditor, which diagnoses site-wide issues; this skill prescribes the full delivery plan for one asset.
-
Formats a raw bug, task, or chore into a clean, ready-to-paste Jira or Azure DevOps work item with title, repro steps, environment, acceptance criteria, severity, and labels. Use when an engineer, QA, or support person needs to file a single tracker ticket fast and consistently.
-
Builds a focused meeting agenda from a goal and attendee list: outcomes, timeboxed topics, pre-reads, decisions needed, and a parking-lot, for kickoffs, standups, design reviews, retros, and steering committees. Use when planning a single meeting (the inverse of meeting-to-action).
-
Writes or audits content against RBA Consulting's brand voice. Apply on any task mentioning RBA, RBA Consulting, or rbaconsulting.com; any RBA named service (AI, Software Engineering, Digital Experience, Data/Analytics, Digital Workplace, Digital Strategy, User Experience, Cloud Infrastructure, Modern Operations); any approved tagline (Powered by People. Enhanced by AI.; Designed to Deliver.; meet you where you are); or any request to draft or audit RBA marketing, blog, case study, email, proposal, LinkedIn, or careers copy. Enforces four voice pillars (confident, human-centered, clear, results-oriented), banned-word swaps (world-class, cutting-edge, synergy, leverage, robust, seamless, empower, holistic, operationalize), em-dash and en-dash with no surrounding spaces, Oxford comma, active voice, no hedging, probabilistic compliance language, inclusive-language replacements, and a self-check before final output. Skip content unrelated to RBA or when a different brand voice is requested.
-
Drafts RBA Consulting proposals for CMS/DXP platform selection RFP preparation (also called pre-envision). Apply when the user asks to draft, scope, or estimate an engagement helping a client prepare an RFP for a CMS, DXP, or similar enterprise content platform — covering discovery interviews, weighted requirements, content strategy and IA advisory, vendor shortlist refinement, conditional uplifts (locations, multi-brand, multilingual, AI governance, data maturity, plus new patterns identified at intake), and recommended RFP structure. Triggers — "draft an RFP prep proposal for [client]", "pre-envision proposal", "CMS evaluation proposal", "envision proposal" for platform selection, "help [client] prepare their RFP", or any RBA scoping where the deliverable is a requirements package for the client's RFP. Apply aggressively — if the conversation references RBA plus client-side CMS/DXP selection, the skill applies even without the word "proposal". Composes with rba-brand-voice for the final voice pass.
-
Builds a 301/302 redirect map from old and new URL lists, matching by slug/path similarity, flagging orphans and chains, and emitting platform-ready output (web.config rules, Sitecore redirect items, Umbraco redirects, IIS rewrite, or CSV). Use during a site migration or replatform when mapping legacy URLs to new ones.
-
Builds, explains, or debugs a regular expression from a plain-language spec or sample data, with a breakdown of each token, edge cases, and ready test strings. Use when an engineer or analyst needs one correct regex for validation, parsing, or a rewrite rule.
-
Walk through a release-readiness checklist before cutting a tag — uncommitted work, open PRs, failing checks, version bump, changelog, and a smoke test of the build artifact. Use before any production deployment.
-
Writes a single, well-formed RAID log entry (Risk, Assumption, Issue, or Dependency) with likelihood, impact, score, owner, mitigation/response, and trigger, in RBA's RAID format. Use when logging one risk or issue during delivery, not generating a status report.
-
Generates a robots.txt and XML sitemap (or sitemap index) for a site, with crawl directives, host/sitemap declarations, and validation against common SEO mistakes. Use when standing up or fixing crawl directives for one site.
-
Explains a SQL query in plain language, reads its execution plan, and recommends indexing or rewrite improvements for SQL Server / Azure SQL. Use when an engineer needs to understand or tune a specific query without a full data review.
-
Authors and validates JSON-LD schema.org structured data (Organization, Article, Product, FAQ, BreadcrumbList, LocalBusiness) for a specific page type. Use when adding or fixing rich-result markup on one template. Distinct from seo-geo-optimizer (strategy) and analytics-tag-reviewer (tracking) — this skill owns the per-template structured data deliverable end-to-end: draft, validate, inject, and hand off.
-
Generates realistic, schema-valid synthetic test data (CSV/JSON/SQL inserts) for a given content model or table, with edge cases, locale variants, and no real PII. Use when seeding a CMS, database, or test environment with sample content or records. Distinct from qa-test-generator (which writes test assertions) — this skill produces the seed data those tests run against. Trigger phrases: "generate test data for", "seed this CMS", "give me fixture data for", "create sample records for", "populate this table with test rows", "generate locale variants for".
-
Turns user-story acceptance criteria into a structured, client-executable UAT test plan — happy-path, edge, error, and accessibility cases with a summary tracker and defect log — exported as a multi-sheet xlsx. Use when a BA, QE, or PM needs formal client-facing User Acceptance Testing cases that a non-technical client can run to validate that what RBA built matches what was scoped. Distinct from automated QA: these are manual, business-language cases owned by the client, not engineering test suites.
-
Turns a feature request, requirement, or rough note into one or more well-formed user stories with INVEST-checked acceptance criteria in Gherkin Given/When/Then format. Use when a BA or PM needs to draft or refine individual backlog stories — not decompose a whole epic (use the epic-decomposition plugin for that). Works across CMS, DXP, portal, and commerce platforms.
-
Authors a single Architecture Decision Record (ADR) in MADR format capturing one decision, its context, considered options, and consequences. Use when an engineer or architect needs to record a discrete technical decision — CMS choice, rendering strategy, auth approach, adapter boundary, IaC tool, caching layer — without invoking a full architecture-advisory session. Supply the decision as the argument; the skill reads the surrounding codebase for context and writes a complete, acceptance-ready MADR file.