Recently updated open-source projects

riff connects Claude Code with OpenAI’s Codex CLI as a peer collaborator rather than a subordinate tool. Its delegate, consult, and roundtable modes can use Codex MCP or `codex exec`, while a trace contract records context, constraints, execution, and acceptance criteria for non-trivial calls. Codex CLI is required; the MCP server is recommended for cleaner multi-turn exchanges.

Codex Skills Library packages common engineering workflows into small, reusable skills instead of requiring teams to repeat prompts. Its domains include codebase orientation, debugging, refactoring, tests, documentation, APIs, frontend and backend work, data, infrastructure, security, performance, reliability, architecture, and planning. Each skill is triggered through natural language and reveals its workflow only when relevant. Users browse the `skills/` directory, select the task they need, and can adapt or extend the library for their team.

iphone-malware-assessment pulls crash logs, creates a resumable encrypted backup, and scans the backup and crash reports with MVT, libimobiledevice, and downloaded indicator feeds. Its report separates actual indicator matches from heuristic notes and includes coverage plus a full file sweep. It runs locally on macOS but cannot prove an unknown zero-day is absent; you need Homebrew, a USB connection, and a backup password you can retain.

tableau-semantics-claudecode-skill packages Tableau Semantics Marketplace workflows as a shareable Python CLI. It runs health checks and produces performance, fleet, revenue, or utilization reports in Markdown or JSON from mock, live Tableau, or hybrid sources. Mock mode works without a live connection; real Tableau access requires administrator configuration, and failed requests intentionally fall back to preserve the workflow.

claude-todos-skill gives each project a committed todo board and SQLite project brain, then aggregates them in a local broadsheet-style Bun dashboard. Its command set covers initialization, batch planning, backend-first autonomous builds, frontend briefs, design, synchronization, QA handoffs, and closing work. Phases, dependencies, history, and evaluation artifacts preserve context across sessions and projects.

Point Mirrorframe at a live URL or local page to capture styles, assets, motion, and interaction state machines, then generate a self-contained React and CSS reconstruction. It replays viewports and interaction states, compares pixels per node, and records explicit failure or skip reasons instead of hiding gaps. Use it only on pages you own or may legally copy; drag interactions, complex physics, and live canvas rendering remain outside full reconstruction.

Docker MCP Gateway is a Docker CLI plugin that can run with Docker Desktop or independently. It isolates MCP servers in containers, organizes them with profiles and OCI catalogs, and gives clients such as VS Code, Cursor, and Claude Desktop one gateway for tools, prompts, and resources. Secrets, OAuth, discovery, and call tracing are supported; Docker Desktop users need version 4.59 or newer with the MCP Toolkit.

SlopCheck flags recurring AI-style habits such as fake profundity, self-narration, forced reframes, and mic-drop endings. Its Guide, Detect, and Rewrite modes can prevent those patterns, score a draft, or return plainer English with a brief note on what changed; portable rules work in other tools too.

Molnify App Builder turns spreadsheet formulas and colored cells into inputs, outputs, charts, and actions for a Molnify web app. It can build from scratch, convert an existing Excel or Google Sheets model, validate the result, and apply styling for calculators, quotes, ROI tools, dashboards, forms, and internal utilities. The open Agent Skill works across Claude Code, Codex, Cursor, Gemini CLI, and other compatible agents, with Python tools for inspecting and validating XLSX files.

substack-migration-check inspects a Substack export ZIP without modifying the original. It creates an integrity and dependency report, cleaned and excluded subscriber CSV files, and a post inventory for hosted images, links, embeds, and paywalls. Processing stays offline: a local script handles subscriber data, while the model sees only aggregate counts and redacted samples.

Positioning Products diagnoses the market story before it writes public-facing language. It interviews you when facts are thin, examines customers, substitutes, proof, timing, and tradeoffs, and can check competitors' actual copy before claiming differentiation. Use it to create a homepage, launch post, name, bio, or pitch, or to critique an existing draft line by line without inventing unsupported claims.

Calibrate the skill with an exported Elementor page, then reuse the site's real widgets, spacing, and global design tokens to generate new `_elementor_data` JSON. It injects pages through WP-CLI and supports single-page or batch builds, with a staging-first draft, verification, and promotion workflow; the target site must be self-hosted WordPress with Elementor and WP-CLI access.

Read variables, dimensions, and structure from a Figma node into a specification table mapped to the repository's design tokens, then inspect the values actually rendered by a web or React Native app. The verification skill distinguishes wrong-token drift, hardcoded values, state variants, and missing elements, with a roughly three-iteration correction loop. It requires Figma MCP plus tooling that can evaluate code in the running app.

screencli-skill lets coding agents such as Claude Code, Cursor, and Windsurf drive a browser from a natural-language task and record the session. The resulting demo can include gradient backgrounds, automatic zoom, click highlights, and cursor trails before being uploaded to a shareable link. It is designed for product walkthroughs and requires Node.js 18 or newer plus FFmpeg.

This skill guides accessible charts in R and Python with color-vision-safe palettes, redundant shapes or line styles, direct labels, contrast checks, and simulations for common color vision deficiencies. Its references cover popular plotting libraries, while a script checks WCAG non-text contrast and CIELAB color distance.

Gausian Native Editor is a cross-platform native video editor for local AI-video workflows. It provides GPU-accelerated preview, asset management, timeline trimming, proxy generation, project persistence, and exports for FCPXML, FCP7 XML, EDL, and JSON. It can automatically import output from a locally running ComfyUI instance, while macOS can optionally open that instance in an embedded WebView. The project is primarily run from source, requires Rust and FFmpeg, uses GStreamer for some proxy and decoding paths, and does not support remote ComfyUI servers.

This Claude skill prepares a Vietnam software copyright dossier for web or mobile projects. It scopes pages, owners, and authors, captures interfaces with Playwright, redacts sensitive data and third-party marks, renders source-code images, and assembles Markdown, DOCX, and paginated PDF files. It is a workflow aid rather than legal advice, so current filing requirements must still be checked with Vietnam’s copyright authorities.

Skillreaper inspects local configurations and session transcripts to show which skills, MCP servers, and agents load without firing, along with estimated context cost. It starts read-only, assigns REAP, MUTE, KEEP, or REVIEW verdicts, and makes pruning reversible through quarantined files and manifests. The tool is local with no telemetry or uploads, but platforms without readable history remain conservatively marked REVIEW.