Shape agent capabilities into your workflow.

This Claude Code skill detects Russian bureaucratic phrasing, English calques, inflated claims, chatbot artifacts, structural habits, and typography patterns. It offers modes for social posts, business letters, personal messages, copywriting, and AI-style audits, with optional voice calibration from sample texts.

Web Design Engineer Pro guides non-technical users from a business description to a finished site or landing page. It generates several standalone browser previews, lets the user choose a direction, then collects a short brief, sets up a Next.js project, builds the sections and motion, and runs build, accessibility, SEO, and performance checks. It requires Node.js 20.9 or newer and can also extract content from an existing site for redesign work.

Turn mathematical or scientific material into complete LaTeX documents, reusable snippets, or Beamer slide decks. After installing the skill, describe the theorem, proof, algorithm, table, or presentation you need, then compile the resulting .tex file in Overleaf or a local TeX distribution. The Claude web route does not require a terminal.

Polish Korean LLM output by targeting stiff formality, empty intensifiers, filler connectives, template openings, hidden agents, excessive hedging, and other catalogued patterns while preserving the text’s meaning. The repository includes a prompt and skill for Claude Code and Codex, plus a shorter prompt for other LLMs.

Use pixeltamer from Claude Code, Codex CLI, OpenCode, Cursor, or another skill-aware agent to generate, edit, or compose images with gpt-image-2. It structures prompts, selects an available backend, and reloads each result for a visual check before reporting success. Authentication can use an OpenAI API key or a signed-in Codex client. Mask-based inpainting and parallel variant generation require the API backend, while standard editing and composition can use either backend.

This Claude Code skill reviews an undergraduate thesis through six weighted dimensions: topic and workload, methodology, data and results, discussion and argument, writing and structure, and references and academic practice. It scans the document, confirms checkpoints, performs chapter-level review by discipline, and produces a seven-part report. Selectable-text PDF, DOC, or LaTeX files are supported; scanned pages need OCR. The score is calibrated for undergraduate work and does not replace a supervisor, formal plagiarism check, or graduate-level review.

MarkPDF Skill lets coding agents such as Claude Code, Cursor, and Copilot render Markdown as themed PDFs, editable PPTX decks, or styled HTML. It can also create share links protected by an email gate, password, or expiry. Rendering runs through the hosted MarkPDF MCP service, so no local document toolchain is needed, but users must configure a service token and the trial tier limits daily exports.

claude-code-deepseek-3p-parity-skill inspects Claude and MCP settings, permissions, proxies, and runtime versions in dry-run mode, then proposes configuration changes for DeepSeek or other compatible gateways. After confirmation it can add model slots, permission rules, trusted workspaces, and MCP entries while backing up JSON and redacting secrets. It cannot grant first-party model capabilities, bypass macOS security prompts, or fix provider limitations.

awwwards-3d gives Claude Code four starting templates for scroll-driven WebGL sites, backed by references for cameras, materials, shaders, post-processing, transitions, and Blender workflows. It can scaffold either a single-file HTML deliverable or a Vite project, and includes a script for exporting compressed GLB assets. The bundled Three.js, GSAP, and Lenis versions are intentionally pinned, so upgrades require compatibility testing.

paper-review.skill guides Claude through a four-stage review of an ML or AI paper: map recent work, test the claimed novelty, write the review, and verify the remaining references. It pushes weaknesses to cite the paper's own evidence and keeps genuine author questions separate from revision requests. It is useful for conference review, pre-submission red teaming, and rebuttal preparation, but is less suited to pure theory, non-English papers, or very narrow fields.

GCC Market Intelligence helps B2B and B2G founders prepare for entry into Saudi Arabia, the UAE, Qatar, Bahrain, Kuwait, or Oman. After narrowing the country, buyer type, and company stage, it surfaces relevant procurement mechanics, localization rules, sovereign funds, soft-landing programs, events, and customer-development frameworks. It is not a live database or legal and tax advice, so changing regulations, quotas, and program terms still need current verification.

This Claude Code skill supports real estate and private-equity asset management after acquisition. It covers quarterly reviews, LP reporting, rent rolls, T-12 and NOI variance analysis, valuations, debt and covenant monitoring, and hold, sell, or refinance decisions across several asset classes. Helper scripts calculate returns, waterfalls, debt metrics, yield maintenance, and operating bridges, while Excel, Word, and slide outputs follow institutional conventions.

This repository packages two Claude Code plugins for Fused. The primary agent-core plugin covers setup, infrastructure, project authoring, execution, verification, storage, and widgets; workbench preserves legacy canvas, JSON-UI, UDF, and integration skills. Both rely on the fused CLI, with the vector extra recommended for local DataFrame handling. It is aimed at developers building and operating Fused projects from Claude Code.

Run a compliance-focused check of a Claude Code setup across network egress, Anthropic routing, long-running automation, sensitive-data exposure, and audit controls. The skill separates locally verified evidence from account, region, payment, and identity facts that only the user can confirm, then returns a risk level and practical remediation steps. It is not a ban-evasion tool and does not advise proxy rotation, fake identities, or unsupported-region bypasses.

Cairn records trust ratings for web pages, search results, MCP tools, and network-bound commands so later agents can check how those resources performed for other reviewers. In Claude Code, post-tool hooks can rate calls in the background; Desktop and web installs expose explicit scoring tools instead. Each rating consumes either Claude subscription quota or Anthropic API spend, and rating data is sent to the Cairn service, so cadence and privacy controls matter.

droit-francais-skill gives Claude a nine-step method for French legal research and requires statutes, case identifiers, and other official references to come from primary sources such as Légifrance. When a citation cannot be checked, the skill marks it unverified instead of filling the gap. Practitioner profiles adapt the workflow for lawyers, in-house counsel, law enforcement, public officials, or students; an optional free PISTE key makes retrieval more deterministic.

Use this skill to turn AI-sounding Italian into natural prose without flattening the author's facts, argument, register, or voice. It applies an Italian-specific editorial workflow covering syntax, punctuation, cohesion, rhetoric, clichés, and common generative habits, and it can also answer usage questions. The skill knows when informal keyboard conventions should be left alone; its instructional content is in Italian.

This skill creates a clinical baseline Table 1 from XLSX or CSV data using design-aware rules: baseline p-values are hidden for RCTs by default, while observational studies emphasize SMD and multiple-testing warnings. It produces self-contained HTML plus Word and LaTeX files, masks identifier columns, and excludes survival, regression, and mixed-effects analyses.