web-design

AI multimedia for brand & winery sites: a creative brief -> on-brand, web-optimized, license-clean images/video/3D/audio behind a mandatory human curation gate. 4 agents (generation-strategist, web-asset-pipeline-engineer, asset-provenance-guardian, brand-and-accessibility-reviewer) route a brief to the right generator (provider-neutral, Grok-lean for images where competitive; inpaint/outpaint/bg-removal/upscale first-class), turn raw output into AVIF/WebP responsive <picture> markup with LCP/CLS-safe embeds, pin commercial-use licenses (flags the FLUX-dev non-commercial trap; C2PA + a provenance ledger; EU AI Act Art.50 disclosure), and gate every asset on brand-hex/style conformance + WCAG alt text before ship. 6 skills, 4 knowledge docs (Mermaid decision trees; all prices [unverified]), 6 best-practices, 4 commands, 5 templates, 4 scripts. The shared foundation the brand-identity-studio plugin consumes. Declarative fal MCP binding (set FAL_KEY). Requires ravenclaude-core@>=0.7.0.

Corporate finance & FP&A specialist team — FP&A analyst, financial modeler, controller, treasury analyst, valuation analyst, audit-prep specialist, and board-pack composer. Ships 23 skills including the controller-autopilot: a governed close-to-report cycle (GAAP statements, COA mapping, reconciliation auto-match, review→approve→lock workflow, ELT staging, consolidation, per-entity dashboard, close schedules) plus its live-integration tier (multi-currency remeasurement + CTA, OAuth GL connectors + drill-through lineage, warehouse/RLS dashboard, IdP-backed segregation of duties) and a gold-standard NetSuite close (OAuth2 M2M + SuiteQL BS/IS trial balance, COA-draft, tie-out doctor, changed-after-sign-off drift, layperson runbook). 10 templates, a knowledge bank, and 2 advisory hooks (anti-patterns + secrets/PII scan). Inherits ravenclaude-core protocols (Grounding, Structured Output, Cited-Adjudicator).

Power Platform specialist team — 11 agents (incl. power-platform-tester, power-bi-engineer) and 23 skills, with strong ALM/git coverage for solutions, flows, and PBIP. A house-opinions hook flags 8 §3/§4 violations; the knowledge bank carries production decision trees (PA-flow recovery, Dataverse token-acquisition, PCF React surface, PBI deploy/refresh, custom-connector build-path, PBIR Enhanced infinite-spinner debug + full build reference, DAX silent-zero scoring via the `Domain` pattern, sempy.fabric notebook reference, Power BI Copilot report-readiness, Code Apps connector gotchas, PBIP deployment variables + #839, PBIR reference enrichment, PBIP report fast-solve triage router [MCP-optional]) plus a real-engagement scenarios bank. Bundles the community pbix-mcp server (d0nk3yhm/pbix-mcp, MIT) for .pbix/.pbit read/write/DAX-eval (`pip install pbix-mcp`); documents (not bundles) the official Microsoft Dataverse MCP (CLAUDE.md §9a). Extends ravenclaude-core.

Project & delivery management team — four specialists across the predictive (PMBOK/PMP) and agile (Scrum/Kanban) tracks and the hybrid between: delivery-lead (charter, schedule, scope/change control, earned value), scrum-master (backlog, sprints, ceremonies, velocity, impediments), risk-and-raid-analyst (scored qual+quant risk, RAID depth, mitigation/contingency, issue triage), and stakeholder-comms-lead (stakeholder register, comms plan, status/exec reporting, escalation memos, steering packs). Deepens — does NOT replace — ravenclaude-core's domain-neutral project-manager (the lightweight RAID/status-hygiene default every plugin routes to); this is the deep PM craft layer. Knowledge: a predictive-vs-agile-vs-hybrid decision tree + a best-practices library. Seams: prose polish → ravenclaude-core/documentarian; system design → architect. Requires ravenclaude-core@>=0.7.0.

AI/LLM red-teaming team — 2 agents (ai-redteam-lead, adversarial-testing-engineer) for the layer answering 'can this AI system be made to do harm, leak data, or exceed its authority — and how do we harden it?': threat modeling + rules of engagement, the attack taxonomy (OWASP LLM Top 10 2025 + MITRE ATLAS), direct vs indirect prompt injection, jailbreaks (roleplay/encoding/many-shot/crescendo), data exfiltration & training-data extraction, agentic tool-abuse / excessive agency, multimodal attacks, and defense-in-depth remediation. Fluent in automated red-team harnesses (PyRIT, Garak, Promptfoo red-team, Giskard) and likelihood×impact severity. 3 skills, a 2-doc knowledge bank (attack-taxonomy decision tree + 2026 patterns), and 2 templates. Distinct from llm-evaluation-engineering (quality-regression eval), trust-and-safety (content-moderation / T&S policy), and security-engineering (app/infra pentest) — the adversarial AI-security layer over model- and agent-based systems. Requires ravenclaude-core@>=0.7.0.