人気上昇中の Agent Skills

Execute Python code in a safe sandboxed environment via [inference.sh](https://inference.sh).

The caveman-stats skill shows real token usage and estimated savings for the current session, reading directly from the Claude Code session log and triggered by /caveman-stats.

Caveman at maximum compression: fragments, one word when one word is enough, each fact once.

Review a Caveman learn report, apply cost-lowering fixes with per-edit consent, and report savings with attribution.

caveman-explore is a read-only repository explorer that quickly locates code and returns compact path:line citations.

Landing page conversion optimization with layout rules, hero section design, and CTA psychology.

AI product photography with studio lighting, lifestyle shots, and packshot conventions.

Film and video storyboarding with shot vocabulary, continuity rules, and panel layout.

YouTube thumbnail design with specific dimensions, contrast rules, and mobile preview optimization.

App Store and Google Play screenshot creation with exact platform specs.

Video ad creation with exact platform-specific specs for TikTok, Instagram, YouTube, Facebook, LinkedIn.

Product Hunt launch optimization with specific specs, timing, and gallery strategy.

Character consistency across AI-generated images with reference sheets and LoRA techniques.

Still-to-video conversion guide: model selection, motion prompting, and camera movement.

Structured competitive analysis with feature matrices, SWOT, positioning maps, and UX review.

Before claiming work is complete, fixed, or passing, you must run verification commands and confirm output; evidence before assertions always.

Execute Python code in a safe sandboxed environment via [inference.sh](https://inference.sh).

Safe refactor skill: restructure code while preserving behavior, guided by verification boundaries.

Turn Caveman's report-only repository observations into an operator-chosen optimization candidate with a paired baseline/candidate evaluation.

Web search and content extraction with Tavily and Exa via inference.sh CLI.

Execute Python code in a safe sandboxed environment via [inference.sh](https://inference.sh).

Render videos from React/Remotion component code via inference.sh.

Diagnose ambiguous failures before editing, rank hypotheses by evidence, and report cause with proof.

Caveman in Classical Chinese: 文言文 register, far fewer characters, technical terms verbatim.

Fix bugs and small behavior changes at the narrowest responsible layer, ensuring regression proof and relevant tests pass.

Prove existing work meets acceptance conditions without expanding scope, for validation-only tasks, completion checks, focused gate runs, and last-mile proof.

Deploy and manage projects on Vercel using the Vercel CLI with token-based authentication.

Web search and content extraction with Tavily and Exa via inference.sh CLI.

Build feature work leanly, avoiding overengineering with a clear scope and stop condition.

Render videos from React/Remotion component code via inference.sh.

Implement reversible compatibility-safe transitions for schema, data, API, protocol, configuration, or dependency migrations requiring rollback and preservation proof.

Discover every LLM workflow in the current repository and label it, so Caveman Cloud groups spend by what the code actually does.

Review Caveman Cloud evidence read-only: costs, Cave Score, Cave Plan, workflows, traces, latency, errors, compression, and verified savings.

Wire the current repository through the Caveman Cloud gateway to measure every LLM request — cost, tokens, latency — with zero behavior change.

Inspect Caveman Cloud's eval-gated experiment lifecycle and block unsafe execution.

Automate browser interactions, test web pages, and work with Playwright tests via a command-line interface.

Audit a landing page, sales page, or checkout page for conversion leaks and return a fix list ordered by expected revenue impact.

Web search and content extraction with Tavily and Exa via inference.sh CLI.

Render videos from React/Remotion component code via inference.sh.

Execute Python code in a safe sandboxed environment via [inference.sh](https://inference.sh).

Turn an offer into a concrete multi-step sales funnel spec - page-by-page structure, price ladder, copy outline and the metrics each step must hit.

Execute a written implementation plan in a separate session with review checkpoints.

Choose a landing page or sales funnel platform by working out the real total cost and lock-in for a specific case.

Give an AI agent its own real browser over MCP tool calls - launch, navigate, click, fill, screenshot, extract text, run JS - with a kernel-level real-device fingerprint and a persistent profile.

Create AI avatar and talking head videos via inference.sh CLI.

Design and implement one-click post-purchase upsells and downsells that raise average order value without hurting the main conversion rate.

Set up server-side conversion tracking so purchases are reported accurately to Facebook, TikTok, Google and Bing despite iOS restrictions, ad blockers and cookie loss.

Verify that browser profiles are actually isolated from one another instead of assuming it—confirm each profile's timezone agrees with its own exit IP, that WebRTC exposes only the proxy, that canvas and WebGL hashes stay identical across relaunches of one profile, and that no two profiles share a persona, a cookie jar, or an address.