Shape agent capabilities into your workflow.

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.

claude-skills-deploy bootstraps a repository with Coolify app configuration, Doppler service tokens, and a GitHub Actions workflow that builds an image once, checks staging, and promotes the same tag to production. Its `validate` and `plan` commands are read-only, while `seed` and provisioning are explicit operations. Using it requires a configured Coolify host, Doppler account, GitHub CLI, SSH access, and local credentials.

frontend-skills helps an agent build or refactor React, Next.js, and Tailwind applications by deciding where code should run before choosing components, actions, or route handlers. It covers Next.js 16 caching, composition patterns, semantic Tailwind themes, common UI shapes, accessibility, and high-impact performance mistakes. The skill is aimed at teams that want fast interfaces without making future changes expensive.

claude-recall contains two Claude Code skills. Recall searches local session transcripts and can transplant a feature from another project with path and import adaptations shown before writing; Recall Pro adds a project registry, status dashboard, focus mode, and memory routing. Both depend on the `context-mode` MCP server, and the Pro component has a separate non-commercial license boundary.

Render a Markdown report, specification, or page draft as HTML, then highlight text, draw region boxes, add comments, or test CSS changes in the browser. The tool exports only new annotations as one structured prompt for the next Claude iteration and tracks their status locally across review rounds. Annotation works from a static file with no browser extension, MCP connection, or server; converting Markdown to HTML requires Python with the markdown package.

focus-router is a text-only Claude Code skill for situations where emotion or urgency might distract from the actual request. It directs attention to the primary task, required format, constraints, risks, evidence, uncertainty, and concrete next steps, while using emotional tone only to calibrate communication. It requires no scripts, dependencies, hooks, or tool permissions and is not a standalone model or multi-agent system.

Use this multi-agent toolkit to inspect Search, Performance Max, App, Display, Shopping, and Video campaigns across one or more manager accounts. It covers conversion and Google tag health, keyword and auction research, pacing, anomalies, asset quality, and placement safety, while keeping exclusions, mutations, and campaign creation behind validation and explicit approval gates. Setup requires Python, gcloud, and a Google Ads developer token, with locally enforced 24-hour sessions.

youtube-summary uses `yt-dlp` to collect a YouTube transcript, metadata, and chapters, then writes a structured note with a TL;DR, takeaways, links, and open questions. For slide-heavy videos, an opt-in mode downloads the video, detects distinct slides with `ffmpeg`, reads their text and visuals, and embeds key images beside the prose. Notes can print in chat or save to an Obsidian vault.

This skill turns dense technical or structured material into continuous prose with plain-language explanations, everyday analogies, and a direct voice. It preserves the source’s meaning and order, asks who the reader is before writing, surfaces ambiguity instead of guessing, and avoids list-heavy formatting and dash punctuation.

This Claude skill recasts complex material as a flowing, flirtatious monologue using gender-neutral pet names, light interjections, and suggestive but non-explicit analogies. It preserves the source’s facts, asks about the intended reader, and keeps consent, safety, and user-requested tone changes as hard boundaries.

This skill turns a project idea into a plan with scope, success criteria, work breakdown, milestones, dependencies, estimates, risks, communication cadence, and a dated Markdown file. It keeps following up across the conversation and adapts the plan when scope or deadlines change, wrapped in a playful Asian-mom persona that can be stopped or softened on request.

Tell the skill what food you have, where you live, your budget, kitchen equipment, dietary limits, and current energy level. It returns practical recipes with quantities and timing, or builds weekly plans that reuse leftovers and produce aisle-grouped shopping lists; optional Notion and Google Calendar integrations can store the results. The default profane chef parody can be switched to a normal voice at any time.

This skill reads the actual Claude usage reset time from the Anthropic usage API, adds a five-minute buffer, and schedules a deferred task for when capacity returns. It depends on Claude Code’s schedule skill and a macOS Keychain credential named Claude Code-credentials, so it is tied to that environment.

Bring it a breakup, conflict, grief, spiral, or difficult message, and it responds as a sharp but loyal friend before helping you decide what to do, draft a script, or rehearse a conversation. The skill does not diagnose or present itself as therapy, and it drops the persona for self-harm or immediate-danger situations.

This skill teaches a structured approach to dynamic programming: define the set of candidate solutions and its attribute, then partition by the last point of difference to derive a transition. It includes a reusable analysis template and examples for 0/1 knapsack, complete knapsack, interval DP, and longest common subsequence.