Recently updated open-source projects

Provide a product idea or URL and the skill evaluates it through the five TELOS dimensions, then returns a go, no-go, or conditional-go recommendation. Quick, standard, and deep modes trade speed for research and estimation detail, while the report can cover team needs, costs, risks, milestones, and next steps. It is intended for early product planning and go/no-go decisions.

Describe a visual such as matrix rain, CRT scanlines, plasma noise, or an image converted to ASCII, and have the coding agent produce a single self-contained TSX component. The references cover procedural animation and image, video, or webcam processing, with multiple art styles, dithering algorithms, character sets, and post-processing effects. React is the only external runtime dependency.

RepoMaster turns a natural-language task into a workflow that discovers suitable GitHub repositories, understands how to use them, and coordinates specialized agents to execute the work. It offers web and command-line modes for deep search, general assistance, or repository-focused tasks. A model key is required, while deep search also needs Serper and Jina credentials.

YouTube Clone KMP is a Kotlin Multiplatform reference project for building a YouTube-style client across Android, iOS, web, and desktop. It demonstrates search, playback, channels, playlists, Shorts, subscriptions, and offline downloads. Running the project requires a YouTube Data API v3 key, and download support needs a separately installed yt-dlp binary; it is not affiliated with Google or YouTube.

CruiseFeed Skill teaches an agent to query normalized live cruise data across more than 60 cruise lines. It selects the right endpoint, filters sailings and fares, handles pagination, and reads only fields returned by the API for questions about dates, destinations, ships, prices, and ports. It is a query guide for the CruiseFeed API rather than a booking system.

AgentTrace reads local session logs from Claude Code, Codex CLI, Gemini CLI, Cursor, OpenCode, and other agent tools, then surfaces token usage, estimated cost, elapsed time, failures, latency, context pressure, and anomalies in a terminal UI. It can compare against a local baseline, export JSON, Markdown, or HTML reports, and fail CI on unhealthy runs. The inspection is local-first, with no hosted backend required for prompts, code, or logs.

Codesplain maintains a readable CODESPLAIN.md alongside a project for people who direct AI rather than write or inspect every line of code. It explains what each area does, why it exists, and which consequential decisions shaped it, updating as the project changes. The file gives you a practical reference when an agent claims to have fixed something, added a dependency, or changed storage.

Use the skill in audit, scan, guide, or harden mode to review a backend project. It detects the stack, maps the attack surface, traces untrusted input to dangerous sinks, checks dependencies and configuration, and reports verified findings with file locations, severity, and fix guidance. Its coverage is anchored to OWASP and ASVS, but it is not runtime testing, a professional penetration test, or a compliance program, and it does not edit code without approval.

Ask what is new today and the skill gathers signals from Product Hunt, GitHub Trending, and Hacker News, then weights them against recent reports and project context. It produces a 60-second summary, three to five deeper reads, priority labels, and cross-source patterns in Markdown, with optional PDF output through gstack. Claude Code WebFetch access is required, and a run typically takes a few minutes.

TanStack CLI scaffolds a full-stack TanStack Start app or a Router-only single-page app, with optional add-ons for authentication, databases, deployment, monitoring, and more. Its JSON commands expose libraries, add-on details, and documentation searches to coding agents. Anonymous usage telemetry is enabled by default and can be disabled by command or environment variable.

This Claude Code skill improves a repository's public presentation by rebuilding the README from verified project facts, refining the GitHub description, and assigning a focused set of topics. When explicitly requested, it can initialize and push a new repository; it can also add a LICENSE file when the README declares a license but the file is missing. It does not change application code, and applying GitHub metadata requires an authenticated gh CLI.

security-hardening offers an audit mode for existing code and a guard mode while building a feature. It checks core application-security areas such as authentication, authorization, inputs, secrets, uploads, and data protection, then ties each finding to a concrete file location and fix. The skill is stack-agnostic and uses a default-deny stance: a domain is not called safe until the relevant code has been inspected.

meta-analysis-skill coordinates a protocol, PRISMA-S search ingestion, dual-blind screening, dual extraction, local R statistics, forest and funnel plots, and an APA or PRISMA-2020 manuscript draft. Seven sign-off gates stop at decisions that carry authorship or interpretation. The bundle also checks citations and typesets the manuscript. Its demo uses synthetic data, and the repository is source-available under non-commercial licenses.

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.

ORCH deploys teams of Claude, Codex, Cursor, Pi, or arbitrary command-line workers, turns goals into tasks, and coordinates parallel execution with waiting, retries, reviews, and status tracking. Coding agents work in isolated Git worktrees and branches so changes do not reach the main branch until explicitly approved. Its shell adapter also allows the same orchestration engine to run editorial, analytics, operations, or other CLI-based processes.

atelier scaffolds a personal site or landing page from a chosen style such as Bauhaus, cyberpunk, Excalidraw, ink painting, pixel art, terminal, or micro-graphics. Each style includes its own design system and runnable site skeleton, with no npm, server, or build step; open index.html directly and replace EDIT placeholders. New styles can be added as independent modules.

Liu Xiaoyan English packages a set of public teaching methods for CET-4, CET-6, and Chinese postgraduate English preparation. It can explain grammar and reading in a conversational lesson mode, break down reading, translation, writing, or cloze questions step by step, and create a staged study plan based on the learner’s goal. The project presents itself as an independent compilation of public materials and is not affiliated with the named teacher or her institution.

Claude Book Skill guides a nonfiction project from interview and voice calibration to research, chapter drafting, editing, reader review, and DOCX compilation. Six specialized roles keep the work separated, while the researcher sources claims or flags gaps instead of guessing. You can provide writing samples from files or connected services, and the compiler can reuse formatting from a sample document.