Skills and agents for Claude Code, Codex, Cursor, OpenCode and more — one install, every harness.
Plugin Claude Code
Skills and agents for Claude Code — visual verification, shareable HTML artifacts, arXiv research intake, model routing, deck and theme systems, and specialist advisor agents.
Agent
Highest-intelligence read-only advisor (Fable 5). Consult at commitment boundaries — before an architecture / migration / API-contract / large-refactor decision, or when a problem has resisted two attempts. Returns a verdict, not a survey. Never writes code.
Agent
Brand strategist for identity systems, visual/voice consistency, and positioning. Use when reviewing or defining brand expression across landing pages, decks, and docs — advisory + brand-guideline deliverables.
Agent
Multi-platform content strategist: editorial calendars, campaign copy, brand storytelling, SEO-aware content. Use for blog posts, release announcements, social copy, and content plans.
Agent
Postgres-focused query and schema tuning advisor: EXPLAIN analysis, indexing strategy, N+1 elimination, partitioning, connection pooling. Use for fleet DB work — slow queries, migration index decisions, schema design reviews.
Agent
Developer advocate for DX audits, technical content, community engagement, and adoption narratives. Use for developer-facing announcements, tutorials with adoption framing, DX reviews of onboarding flows, and conference/demo material.
Agent
Fast read-only search/exploration lane (Haiku 4.5). Use for broad fan-out — locate code, map naming conventions, gather files across a large tree, summarize or classify at volume. Returns conclusions, not file dumps. Never writes.
Agent
Hard-build lane (Opus 5) — the write lane for escalated, uncertain, or reasoning-heavy implementation and root-cause debugging. Use when a change is hard as code (gnarly logic, a non-trivial algorithm, a stubborn bug) or when the implementer (Sonnet) lane has failed the same task twice. Writes code and proves it.…
Agent
Identity and access domain advisor: OAuth2/OIDC, SSO, SCIM, WebAuthn, multi-tenant RBAC, session and token design. Use for multi-tenant SaaS authz work — reviewing authorization flows, token lifecycles, or tenant-scoping designs.
Agent
Standard implementation worker (Sonnet 5) — the build lane. Use for normal feature/bugfix implementation and moderate-reasoning multi-file changes where the approach is already decided. Writes code and proves it with tests.
Agent
Advisor for production multi-agent system design: topology selection, context management across agents, inter-agent trust, failure recovery, HITL gating, agent observability. Use when designing or reviewing autonomous agent lanes, orchestration scripts, or agent-team patterns.
Agent
Read-only research analyst — deep-reads ONE arXiv paper's LaTeX source and returns a typed relevance verdict against the current project. Use to evaluate a paper without spending main-session context, or as a fan-out lane under an arxiv-sweep workflow. Cites file:line for every claim about the codebase and never…
Agent
Simulated-user cognitive walkthrough reviewer: steps through a UI/page as defined personas, applying CRO and usability frameworks to find friction. Use to review specs-dashboard views, landing pages, or docs flows against the reading-UX north star.
Agent
Creation-side orchestrator for anything you have to present. Takes "I need to present X to Y (at Z, in N minutes)" and delivers the finished piece — picks the surface (house deck / HTML artifact / pptx / HyperFrames video / slideshow), applies the design-kit theme the context calls for and asks which font pairing…
Agent
Strong read-only reviewer / synthesizer (Opus 5). Use to review a completed diff for correctness, planted frauds, and scope creep, or to synthesize multiple sub-results into one coherent answer. Cites file:line. Returns a VERIFIED / VERIFIED WITH CAVEATS / REFUTED verdict. Never writes code.
Agent
Search relevance advisor for Elasticsearch/OpenSearch: index/analyzer design, BM25 tuning, query DSL, relevance evaluation. Use for fleet OpenSearch work — index mappings, search quality, analyzer choices.
Agent
Statistical rigor advisor: experimental design, signal-vs-noise, pressure-testing metric claims, sample-size and confounder review. Use when designing evals, judging telemetry/dashboard metrics, or validating AI-sprint measurement claims before presenting them.
Agent
Expert technical writer for developer documentation, API references, README files, tutorials, and guides. Use for writing or overhauling docs, landing-page copy with technical depth, or turning complex concepts into clear developer-facing prose.
Agent
Technical UX architect for CSS design systems, responsive layout frameworks, information architecture, and implementation-ready UI specs. Use for dashboard/landing-page structure, design-token systems, and dev-handoff specs.
Command
Pick a Claude Code output style from a popup (attention-span styles).
Skill Claude CodeCodex
Turn arXiv research into a decision for the project you are in — scan a category batch, deep-read one paper's LaTeX source, or map a corpus onto this codebase and produce a scored plan. Enforces a citation gate so no figure is ever paraphrased into existence. Use when: "check recent arXiv papers", "what's new in…
Skill Claude CodeCodex
Claude-native multi-model routing — dispatch each task to the right Claude model tier (Fable judgment · Opus orchestrate/review + hard build · Sonnet build · Haiku volume) via model-pinned subagents. Use when routing a Claude task by model or reasoning effort, consulting the advisor at a commitment boundary, or…
Skill Claude CodeCodex
Route bounded implementation work from Claude to Codex using Luna, Terra, and Sol profiles or an opt-in Sol-led subagent team. Use whenever Claude dispatches, resumes, reviews, or rescues Codex work, or when a Codex handoff is oversized, fails, or needs escalation.
Skill Claude CodeCodex
The shared design layer behind every surface you present on — one token registry (11 themes × 9 font pairings, validated as a pair), a local webfont pipeline with a license manifest, cross-surface motion tokens, and per-theme chart ramps. Compiles to house-decks (drop-in data-theme block), html-artifact (three-layer…
At most 3 mods per repository are shown here — the rest are on their repository pages: