Recently updated open-source projects

Use visual-cook to interview through fuzzy design language and test decisions on the real component rather than a mock. A permanent keeper harness records current and rejected lanes, rationale, settled values, and whether the result reached production, while an index prevents old ideas from being cooked again. The provided skeleton assumes Next.js and Tailwind.

LinkedIn Profile Optimizer audits personal profiles, company Pages, target-profile fit, or a combined founder-and-Page setup. It rewrites sections such as the headline, About, experience, skills, and brand description, then adds relationship planning, content direction, social selling, ads, compliance, and KPIs. The focus is the LinkedIn asset and relationship strategy, not drafting posts or articles.

Countries API Skill helps an agent choose endpoints and parameters for the free countries.dev REST API. It covers country and border data, cities and reverse geocoding, postal codes, IP geolocation, reference data, and great-circle distance. No API key or signup is required, and the repository includes a full endpoint reference for Claude Code or other compatible agents.

AgentVerse offers separate frameworks for collaborative task solving and for simulating interactions among multiple LLM-based agents. Its examples cover software teams, classroom environments, games, and other research scenarios, with support for hosted APIs and selected local-model setups. The simulation code is being refactored, and the project directs users who need a stable simulation-only version to the release-0.1 branch.

auto-support is a Claude Code plugin for answering a product’s Discord users from an allowlisted set of public documents. It uses permissions, a fail-closed tool hook, standard-library detection, and an egress DLP gate so the model cannot open secrets or source files. Out-of-scope or uncertain questions are refused and escalated to founders; the MVP produces drafts for a user to review rather than auto-posting.

paper-figures reads a manuscript and its raw data, chooses statistics and chart types that match the claim, and applies journal presets such as Nature, IEEE, or PLOS. It renders and self-checks the figures, then exports editable vector graphics, high-resolution raster files, three-line Word tables, and an English, Chinese, or bilingual report. Every result comes from runnable analysis code, but authors still need to verify the methods and captions before submission.

Run one query across several free media services and receive a combined list of images, videos, or GIFs labeled with their source and license. web-media-getter can also download selected files and write an attribution.json file beside them, so provenance travels with the assets. Five sources work without keys, while services such as Pexels, Pixabay, and GIPHY require their own free API credentials.

macos-screen-recorder captures the main macOS display and the sound your Mac is playing, then writes a shareable H.264/AAC MP4 from a headless command-line run. It uses ScreenCaptureKit in one Swift source file, needs only the standard Screen Recording permission, and avoids virtual audio drivers or sudo. Duration is set when the command starts, and window or region selection is not included.

lookdev creates a live browser studio where a person tunes visual output with sliders, color pickers, drag handles, crop controls, and inline annotations while the render updates immediately. The chosen state can be written back as reproducible JSON, settings, code, or 3D metadata, so the deliverable is more than a screenshot. It covers image, color, typography, layout, animation, and material work, and also supports prose and media comments.

Render several clearly labeled variants into one contact sheet or motion sequence, then ask a vision or video model to score them and suggest the next parameter values. The Claude Code skill structures a coarse-to-fine loop with calibration anchors and an early-exit rule, reducing manual trial and error. Its judgment is still an opinion, so personal taste and brand choices should remain human decisions.

deterministic-design renders the actual interface before judging it. A spatial audit measures the ink-weighted centroid against the optical center, checks the declared grid, spacing, alignment, contrast, and touch-target size, and draws findings back onto the screenshot. A separate fresh-eyes vision pass evaluates usability against Nielsen’s heuristics and produces a prioritized fix list. The skill composes with other design tools and adds measurement rather than replacing visual taste.

CKW Design packages frontend direction, design-system practice, spatial composition, and visual philosophy into composable Claude Code skills. It helps define a product-specific tone and color world, translate that direction into tokens and typography, and review rendered output instead of accepting the model's first layout. The collection is aimed at web components, dashboards, landing pages, and similar UI work.

production-audit examines a project’s live URL, public GitHub signals, and repository structure to surface high-impact launch risks such as webhook idempotency, RLS gaps, exposed secrets, missing rate limits, and database indexing issues. It complements in-editor security review by checking the state users actually receive after deployment. The service is primarily intended for public repositories with a live app and calls the commit.show audit API.

MockHunter opens a live page through Claude Code and Playwright MCP, exercises its interface, and traces each visible value through network calls, APIs, and an optional database. The report classifies data as real, mock, LLM-generated, hardcoded, broken, or unknown, with severity and recommended actions. It avoids destructive controls and payment submission, but currently audits one page per run and may require manual pre-login for OAuth, magic links, or two-factor authentication.

demand-mining turns a shipped product’s Discord conversations into a daily demand radar. It removes PII before the model sees messages, extracts explicit and implicit jobs-to-be-done, deduplicates across days, and ranks the pool with RICE, Opportunity, WSJF, and Kano views. The result is an end-of-day brainstorm and prioritized iteration queue; search, verification, and listening are delegated to companion tools.

Markdownify is an Electron desktop editor that renders a live HTML preview while you type and keeps the preview aligned with the current editing position. It supports GitHub Flavored Markdown, syntax highlighting, KaTeX, themes, multiple cursors, basic formatting controls, and PDF export, with an optional tray presence for quick access. Installable builds are available for Windows, macOS, and Linux.

daily-hotspots gathers opportunity candidates from multiple sources, merges them across days, applies a reproducible scoring rubric, and delivers tiers through Discord with a private archive. A fail-closed gate requires at least two independent origins and keeps deep dives within a daily limit; deeper research is delegated to market-intel or small-cap-deepdive. It owns cadence and ranking, not the underlying search or verification engines.

Copilot Brag Sheet records files changed, pull requests, and git activity during Copilot CLI sessions, then stores the data locally for later review and work-log generation. It is designed to turn forgotten tasks into evidence-backed impact statements for performance reviews. Automatic session tracking currently depends on Copilot CLI, while other MCP hosts can use the tools only through explicit manual capture.