Find the open-source solution that fits.

This skill routes an image request from Claude on the web or desktop through Desktop Commander MCP to a locally installed Codex CLI, then saves the resulting PNG and displays it back in the conversation. It requires a Mac, an authenticated Codex installation, image generation enabled in Codex configuration, and a working Desktop Commander connection. The workflow is limited to single text-to-image requests; batch generation, image editing, video, and transparent-background output are outside its scope.

Run the skill against a strategic deliverable to receive a PASS, REVISE, or FAIL verdict, scores across eight consulting-oriented dimensions, and a concrete repair directive. Its Skeptic Agent probes for silent ambiguity choices, agreement with flawed premises, and missing counterarguments, while a separate novelty stack checks whether the insight is more than generic strategy language. The evaluation runs inside Claude Code without separate provider API keys.

Configure topics, RSS feeds, brand voice, and target channels, then run the skill to filter relevant items and check keyword or trend signals using the best available source. It writes a daily Markdown file with drafts tailored for LinkedIn, X, blogs, or newsletters. It never publishes automatically, and basic operation requires an RSS or web-fetching tool such as Firecrawl or web_fetch.

This Claude skill grounds analysis of Chinese classical texts in provenance, authorship, editions, biography, historical context, and commentarial traditions before moving to interpretation. It preserves original passages, avoids unsupported attribution and fabricated citations, and postpones modern analogies unless requested. The scope covers classics, histories, philosophy, Buddhist and Daoist scriptures, and Yi studies; plain translation and simple lookup requests are outside its trigger.

Use this Claude Code skill to build single-file HTML pages with a defined type scale, spacing system, one accent color, inline icons, and a light theme with dark-mode toggle. Its validators render mobile, tablet, and desktop screenshots with Playwright and run static and accessibility checks. The skill repo needs Playwright, Sharp, and Chromium installed for the full validation workflow; consumer projects receive the generated page, not copied tooling.

Use libu-studio to track prompts, generation runs, candidates, costs, character profiles, and delivery rules for liblib.art assets in one SQLite database and CLI. The workflow supports icons, images, videos, and WebP animations, including optional injection into a target project. Foreground masking and despill require macOS 14+; generation also needs a liblib account and credits, while other systems are limited to the data model and scripts.

pubmed-reference-resolver extracts References sections from PDF, DOCX, or TXT files, resolves citations against PubMed, and writes a PubMed-compatible CSV, numbered abstract text, integrated audit report, and journal-mismatch JSON sidecar. It handles common citation styles, PDF line-number noise, an offline MDPI fast path, unresolved-reference preservation, composite-key deduplication, and possible fabricated-citation detection. Python 3.11+, Anthropic, and NCBI credentials are required.

Audit a repository, branch diff, or selected directory for code that tells users an action succeeded when it did not, as well as mocks, empty handlers, stale placeholders, and unfinished paths exposed as real features. Two complementary specialists inspect failure paths and apparent happy paths in parallel, then a deterministic aggregator produces a severity-ranked combined report. The silent-failure specialist requires a separate Anthropic plugin, while the stub audit can run alone.

github-repo-audit runs five specialist agents over an existing repository’s public surface, covering competitive framing, user language, positioning, README hierarchy, and search or AI discoverability. It synthesizes the findings into severity-ranked issues and proposed fixes, then waits for approval before editing files or GitHub metadata. The skill is for maintainers with a working product, not for code quality or security scanning.

This skill turns a public Xiaohongshu image-note link into a structured reading with the title, original text, all images, engagement numbers, comments when available, and a one-line takeaway. It can optionally archive Markdown and images locally, but video and private notes are unsupported; comment capture needs Claude in Chrome, a Max subscription, and a logged-in browser.

Inject the self-contained script into a prototype or load it as a bookmarklet, then press M and click any element to attach a note. Pins stay anchored during responsive changes and persist per page in localStorage. Feedback can be copied as Markdown, plain text, JSON, or an AI-oriented handoff containing CSS selectors and HTML snapshots.

This official PGYER skill can be used from Claude Code, Cursor, or Codex to upload ipa, apk, and hap packages, retrieve install links or QR codes, check app versions, and generate CI/CD templates. It prefers the official MCP server when available and falls back to PGYER’s shell and API flow; a PGYER API key and account access are required.

This skill runs a read-only audit of a repository in full, differential, or targeted mode, checking security, bugs, workflow, UX, and types before deduplicating findings into Markdown and JSON reports. It can collect lint, type-check, test, dependency, and Git baselines; audit depth depends on the specialist skills installed.

codex-oi lets a coding agent request an independent review without leaving the terminal. It can audit a file or project, challenge a plan, inspect the current diff before shipping, or review a landed commit, using Codex by default with Gemini available as a plugin engine. It adds project context, enforces filesystem boundaries, and formats findings, while keeping every result advisory and never applying changes automatically.

This reference skill organizes 294 deduplicated Chinese PowerPoint design tips into 14 categories covering layout, color, typography, charts, imagery, animation, presentation logic, workflow, and common pitfalls. It works as a Claude Code skill or a standalone handbook, focusing on quick design guidance rather than generating or rendering decks.

Run static checks across common web frameworks and dynamic browser probes for language, headings, ARIA landmarks, form labels, contrast, focus visibility, and related issues. The plugin maps findings to all 106 RGAA 4.1.2 criteria and can generate an accessibility statement, a three-year accessibility plan, an annual action plan, and a traceable CSV grid. Its automated scan intentionally favors false positives, so expert verification remains necessary before official publication.

This tool fetches the day’s limit-up stocks and analyzes each one across news catalysts, capital participation, limit-up quality, sector resonance, and consecutive-board position. It produces scored Markdown attribution reports with a driver summary and next-day outlook; the project is for research only and does not provide investment advice.

Give a small business one self-hosted, installable home for its internal web tools, with authentication, app management, and shared PDF, email, and per-user storage services. The Claude Code skill can scaffold, package, and upload child PWAs; the claude.ai version can only generate files and instructions. Each deployment is single-tenant, runs with Docker, and needs wildcard DNS for per-app origin isolation.