Recently updated open-source projects

Eagerworks' skills collection shows a portable layout for Agent Skills: each skill has a frontmatter-based SKILL.md, optional references, and installable assets. It works with Claude Code, Cursor, Copilot, Codex, Amp, and other compatible agents through the skills.sh CLI. The current collection contains the Kamal deployment skill, covering deploys, rollbacks, proxy, SSL, builders, and troubleshooting.

This adaptive Claude skill supports Brazilian LGPD work for DPOs, privacy engineers, and executives. It covers data-subject requests, 72-hour incident response, DPIAs, ROPA, vendor agreements, international transfers, consent, and retention, with PT-BR templates, PII/cookie/TLS scanners, and tamper-evident evidence logs. It is compliance tooling, not legal advice.

Use this Claude skill while building a WordPress plugin, addressing Plugin Check warnings, or reviewing a directory closure notice. It covers capabilities, validation and escaping, REST permission callbacks, internationalization, privacy, and the 19 wp.org Plugin Directory guidelines through lazily loaded references. It intentionally excludes themes, standalone PHP, and purely frontend work.

claude-favicon-gen takes one square master image, ideally at least 512px, and writes a multi-resolution favicon.ico, small PNGs, an Apple touch icon, Android Chrome icons, and a site.webmanifest when needed. It also provides the references needed for an HTML head or Next.js metadata.icons configuration. The Claude Code plugin runs locally with ImageMagick, so the logo does not need to be uploaded to a third-party service.

svg2pptx-skill rebuilds an SVG as native PowerPoint DrawingML so shapes, labels, arrows, gradients, and groups remain editable. A compatibility gate catches unsupported SVG features before conversion, and the tool can process one diagram or a folder into a deck that can be exported to PDF, Keynote, or Google Slides.

audit-croatian-gdpr-compliance-skill classifies the product, stage, data subjects, data categories, and controller or processor role before applying the EU GDPR baseline and Croatia-specific gates. Each check is marked pass, fail, risk, not applicable, or needs legal confirmation, with observed basis and article references in a self-contained HTML report. Its source register is dated, so refresh official sources before relying on high-stakes or client-facing conclusions.

This collection helps AI assistants handle common Dart and Flutter work through focused guides for architecture, persistence, adaptive interfaces, and testing. Each module explains when to use it and provides an actionable workflow for implementation, refactoring, or troubleshooting, with additional skills for AGENTS.md generation and backend-driven UI.

Start with recent AI or technology topic ideas, approve an outline, and produce a WeChat Official Account article using one of four structures: trend analysis, explainer, case study, or opinion. The skill also checks for common AI-writing habits and supports standalone research or rewriting tasks. With optional WeChat API credentials, it can save the finished article directly to the draft box.

Pagedraw is a visual UI builder for React that works more like Sketch or Figma than a conventional code editor. It stores the design in a Pagedraw document, generates JSX and CSS into the project, and updates a local React app as you edit. The generated files remain regular source code if you later remove Pagedraw. Development has stopped and the maintainers explicitly advise against production use, so it is mainly useful for experimentation, study, or a fork.

sign-safe decodes an opaque Solana transaction before signing, classifies dangerous instruction shapes from the signer's perspective, and emits a SIGN, HOLD, or REJECT verdict plus machine-readable JSON. Its core requires no key, RPC, or broadcast; optional simulation and RPC enrichment are separate, and SIGN only means the transaction passed the current static policy.

tighten removes filler connectives, throat-clearing, hedging, empty intensifiers, repeated context, and other padding while preserving every fact. Numbers, paths, code blocks, error strings, decisions, and caveats stay intact. It can edit a file or section in place or return a rewrite for pasted text; the goal is tighter prose, not summarization or a different argument.

Joplin Favorites adds a dedicated panel for keeping notebooks, notes, to-dos, tags, and searches within easy reach in the desktop app. Favorites can be renamed, reordered by drag and drop, and displayed in horizontal or vertical layouts. Saved searches are only partially supported: selecting one copies its query and focuses the global search bar, so you still need to paste and run it manually. The plugin requires Joplin 2.1.5 or newer.

QA Skill turns an AI agent into a browser-driven manual-QA assistant: it explores an app, drafts test cases, runs existing cases, and records the results while leaving bug judgment to a person. Explore, run, and update modes cover discovery, execution, and keeping cases aligned with UI changes. Test cases live in a Markdown suite and run history in JSON; ambiguous or unobservable outcomes become Uncertain or Skipped.

Kelly Advisor classifies a decision as repeatable betting, portfolio allocation, or a one-shot irreversible choice. It uses fractional Kelly and downside checks for the first two, but refuses to output a number when the probabilities are unknowable, instead asking survival and fallback questions. It is not a stock-tip or return-guarantee tool.

bug-scanner reviews the current Git diff by default or scans paths you name, looking for null access, missing awaits, races, injection, missing authorization, SSRF, unsafe deserialization, N+1 queries, blocking I/O, and similar defects. Findings are grouped from Critical to Low with file locations and concrete fixes. It only reports issues with a real consequence, not naming, quote-style, or formatting nits.

questionnaire-analysis-skill guides bilingual survey and scale research from questionnaire design and data cleaning through reliability and validity, group tests, regression, SEM, mediation, moderation, and APA reporting. It packages methodological references, result-sentence templates, checklists, and an offline threshold reference. The approach is tool-neutral but centers on SPSS and AMOS, with R examples for key steps.

This Codex and Claude skill reviews changes as if they are about to run in production, prioritizing correctness, data safety, security, reliability, and missing tests over style. Findings include severity, confidence, location, impact, and a suggested fix, while the final verdict is derived from the highest-severity confirmed problem rather than reviewer mood.

motion-mosaic-editing teaches an engine-agnostic method for adding motion graphics, kinetic captions, overlays, and infographics to existing footage. It decomposes an edit into reusable Tiles, keeps an element-addressable JSON scene specification, and moves from intake and analysis through tile design, captioning, reframing, scoring, export, and reuse. The skill explains how an LLM can decide while Remotion, SVG and Chromium, Canvas-WebGL, or FFmpeg perform the rendering. It ships methodology, not a renderer or proprietary Mosaic code.