Recently updated open-source projects

qpay-skills guides Mongolian developers from stack detection through QPay v2 setup, credential configuration, invoice creation, webhook handling, status polling, local testing, and a launch checklist. It includes patterns for Next.js, Node, Python, and Laravel plus bilingual merchant guidance. A registered QPay merchant account, contract, and API credentials are required, and the community project is not an official QPay product.

Marketing Audit Skill gathers audits, SEO analysis, copy and ad drafting, funnel reviews, competitive analysis, and launch planning behind one /marketing dashboard. Fifteen sub-skills use a shared scoring rubric and require recommendations to name the change, location, and expected impact. Lazy loading keeps each run focused, and missing pages or inputs are reported instead of filled with invented analysis.

TradingView Session Skill turns the startup ceremony into a single tv_start trigger that launches TradingView Desktop, checks the local MCP connection, verifies a rules.json knowledge base, and returns a compact readiness status. It has a fallback launch path when the normal MCP action fails. TradingView Desktop, a local companion MCP server, and a machine-specific rules file are required.

frontend-post formats user-supplied Chinese copy into Xiaohongshu-style cards at fixed 1080×1440 or 1080×1080 dimensions. It offers twelve curated visual presets and exposes colors, fonts, decoration, spacing, and copy limits for direct adjustment, while leaving the content itself under the user's control. PNG export uses Chrome with Puppeteer or Chromium with Playwright; it is a layout tool, not a copy generator.

ikuai-router-skill combines an Agent Skill with a standalone Python helper for the iKuai router's private `/Action/login` and `/Action/call` API. It covers 14 domains, including network, DHCP, NAT, routing, QoS, firewall, VPN, wireless, logs, and system tools, with authentication caching and automatic re-login on token expiry. The catalog was verified against iKuai 3.7.8 on an IK-M50; the API is undocumented, firmware behavior may change, and destructive operations can lock you out.

shopping-compare researches purchases across Chinese e-commerce platforms, combining cross-site comparisons, independent reviews, safety records, and checks for misleading marketing claims. It starts by asking whether the purchase is necessary, then gives reasons, alternatives, and risks while leaving the final choice to the shopper. Direct access to several platforms is blocked, style products may require screenshots, and web prices can become outdated.

Travel Agent turns a trip idea into a seven-phase plan: clarify priorities, shape destination options, check live logistics, build a paced day-by-day itinerary, track reservations, prepare packing, and export a local PDF. Each day gets a named anchor and backup, with advice shaped around the traveler's interests. It is not a booking or fare-aggregation service, and its source-available license is non-commercial.

home-router-tune discovers router settings and nearby Wi-Fi conditions, builds a risk-ranked plan, applies only approved changes, and writes a report with history for later drift checks. Execution stops at the first verification mismatch and includes a rollback recipe plus a configuration backup when available. It depends on dev-browser and ships detailed playbooks for several Brazilian ISP gateways, with a generic fallback elsewhere.

macleanse performs a CleanMyMac-style scan, classifies findings by risk, proposes cleanup actions, and reports what changed. Destructive scripts default to dry-run, sizes are shown before deletion, and denylisted locations such as Photos, iCloud Drive, Keychains, and backups are protected. It is macOS-only and requires explicit confirmation before applying cleanup to user data.

oracle-26ai-helper-ptbr provides Brazilian Portuguese guidance and executable SQL templates for Oracle AI Database 26ai features, including AI Vector Search, JSON Relational Duality, and RAFT replication. Its rules emphasize version checks, bind variables, vector-memory sizing, ETag concurrency, and Trivadis naming. The documented target is the 23.26.x Enterprise Edition release; it does not apply to Oracle 19c, and RAFT coverage is exploratory rather than a deployment plan.

Run /report inside a Git repository and the Hootbu Daily Report Skill reads the selected commits, groups related work, and rewrites code-level messages into paragraphs suitable for a manager or client. Turkish is the default, with English available through a flag. After your name, company, and project are saved once, --direkt can append the day's report directly to a Markdown file on the desktop.

Ask a compatible agent to turn Markdown into a formatted Google Doc, place it in a designated Drive folder, enable link-based viewing and commenting, and return the share URL. Previously published documents can be updated without maintaining a separate local state file. The workflow requires the Google Workspace CLI and Drive OAuth access, scans every create or update for common secret patterns, and demands explicit confirmation before overwriting a document outside its managed folder.

The setup pairs two Windows machines with Syncthing, replicates the Claude Code Desktop session folder, and runs a PowerShell merge script at login so conversations from the other workspace appear locally. It does not support the Claude CLI, macOS, or Linux, concurrent offline edits can conflict, and Microsoft Store sandbox redirection must be checked during installation.

Give Claude Code a firm but non-abusive review persona that challenges premature architecture, unjustified rewrites, and other red flags before helping. Nonna can flag technical debt, missing tests, and oversized files, respond in Italian or English, and switch between Full, Lite, and Del Sud modes. Hooks and slash commands are Claude Code-specific; the skill itself is offline-first with no external runtime dependencies.

This Claude skill guides short- and long-form novel creation through 13 Snowflake Method steps, from theme and one-sentence premise to characters, layered outlines, scene planning, chapter generation, and prose conversion. Its final pass scans for technical wording, awkward logic, and stiff tone, then applies five style adjustments and shows the original alongside the revised text.

Context Window Management adds operational rules for long Claude Code runs: hard caps on subagent handoffs, a marker-line protocol that accounts for injection and truncation, a launch-turn barrier before merging results, and eight named anti-patterns. Use it before kickoff only when at least two conditions hold: raw tool output is large, several agents reuse the same artifact, or three or more parallel branches need to be merged. If only one condition holds, the project recommends a flat plan instead.

The skill searches 20,972 Arabic-English records from the Quran, Sahih al-Bukhari, and Sahih Muslim through Supabase and pgvector. It supports exact lookup, topical search, cross-referencing, and checking whether a report appears in the two Sahih collections, while presenting returned text with citations instead of reconstructing it from memory. It does not issue fatwas and requires a Supabase deployment plus embedding setup.

harness-engineering-skill bootstraps a project harness for reliable LLM-agent work and maintains it across sessions. Its four modes cover project setup, handoff, promoting repeated mistakes into stronger guardrails, and independent evaluation. Deterministic shell scripts handle repeatable operations, while Markdown modes leave judgment to the agent; project artifacts record configuration, progress, features, lessons, and evaluation reports.