blacktop/dotfiles
14 skills274 installs238 starsGitHub
Install
npx skills add https://github.com/blacktop/dotfilesSkills in this repo
1Ratatui TuiRatatui-tui helps build terminal user interfaces with the ratatui Rust crate. A developer uses it when creating new TUI apps, adding widgets and layouts, wiring keyboard navigation and state management, or integrating images and loading animations. It covers the v0.30.1 API, the Elm Architecture, StatefulWidget, async event handling, and release optimization, with copyable templates to start from.247installs2Code Simplifiercode-simplifier is a Claude Code skill that refines recently modified code for clarity, consistency, and maintainability without changing what it does. It applies project standards from AGENTS.md, reduces unnecessary complexity and nesting, improves naming, and avoids over-simplification like nested ternaries that hurt readability. A developer relies on it to clean up freshly written code, and it operates on the recently touched code unless told to review a broader scope.5installs3Rust ProfilingRust-profiling helps profile Rust binaries with samply to identify CPU bottlenecks. A developer uses it when performance is slower than expected, before optimizing to measure first, or after optimization to verify improvement. It covers building with debug symbols, recording with samply into the Firefox Profiler UI, and analyzing saved profile.json files via a bundled script.5installs4Are You Sureare-you-sure is a self-review skill that has an agent do one careful second pass over the code, config, tests, or docs it just changed. It reconstructs the exact scope, re-reads the diff, checks five risk buckets, and fixes clear local issues before finalizing, then reports what it fixed and what remains risky. A developer uses it as a quick self-check before committing, opening a PR, or handing off, and it explicitly is not a substitute for a real security audit or external review.3installs5Handoffhandoff is a Claude Code skill that generates a handoff prompt another LLM agent can execute without guessing. It picks a shared-workspace or fresh-context mode, gathers only execution-critical context like objective, success criteria, scope, files, and verification steps, fills a labeled template, and applies tuning notes for the target model family. A developer uses it to delegate a task to another agent or start a fresh session on a different model with a clean, atomic brief.3installs6HumanizerHumanizer edits or reviews text to remove signs of AI-generated writing and make it sound more natural. A developer uses it when cleaning up LLM-generated first drafts or reviewing content before publishing. It detects patterns like inflated symbolism, promotional language, em dash overuse, and the rule of three, drawing on Wikipedia's 'Signs of AI writing' guide.2installs7Screenshot AnalyzerScreenshot-analyzer finds recent screenshots and analyzes them for UI/UX review. A developer uses it to inspect interface errors, review layout issues, or examine visual problems in a screen capture. It reads the image and evaluates it against a checklist covering visual hierarchy, layout and spacing, typography, color and contrast, interactive elements, and error states, then returns structured findings.2installs8Apple Native Devapple-native-dev is a Claude Code skill for building, signing, and deploying iOS and macOS apps using CLI-only workflows and the Zed Editor instead of the Xcode UI. It covers generating projects with XcodeGen, managing credentials with xcconfig files, wiring up the Swift LSP in Zed, building and deploying to simulators or physical devices via justfile commands, and scaffolding GitHub Actions CI. A developer uses it to set up or run an Apple app project without depending on Xcode's interface.1installs9Go Performancego-performance is a Claude Code skill for measuring and improving Go program performance using a current Go 1.26-era workflow. It insists on measurement before rewriting: name the metric that matters, add or repair a benchmark, compare runs with benchstat, collect one profile at a time, fix the dominant cost, then re-measure. A developer uses it when profiling Go code, diagnosing CPU or memory bottlenecks, writing benchmarks, or applying PGO to hot paths.1installs10Instruments ProfilingInstruments-profiling helps profile native macOS or iOS apps using Instruments and the xctrace CLI. A developer uses it to record a Time Profiler trace by launching or attaching to a process, then open or export the trace for stack analysis. It covers correct binary selection, CLI arguments, and common gotchas like LaunchServices resolving the wrong bundle.1installs11OrchestrateOrchestrate helps split a task into scoped sub-tasks and delegate bounded work to Claude Code via the claude -p CLI while Codex stays lead developer. A developer uses it to run parallel Claude sessions, isolate risky edits in worktrees, define custom subagents, or move repeated enforcement into hooks. It keeps architecture, acceptance, and integration decisions local and delegates only mechanical or bounded work.1installs12Save SessionSave-session saves AI agent session information so any agent can resume the work later. A developer uses it when finishing a work session and wanting to save progress, or before ending a long coding or debugging session. It writes session metadata (session ID, agent, summary) to docs/.ai/sessions.json via a bundled script and documents how to find the session ID for Claude, Codex, Gemini, Copilot, and Cursor.1installs13Seatbelt SandboxerSeatbelt-sandboxer generates minimal, allowlist-based macOS Seatbelt sandbox configurations for applications. A developer uses it to sandbox, isolate, or restrict a macOS process that needs limited file and network access, or for defense-in-depth when supply-chain attacks are a concern. It profiles an app's requirements across categories like file read/write, network, process, and Mach IPC, then builds a deny-all .sb profile that allows only what is needed.1installs14Second OpinionSecond-opinion runs an independent, external LLM code review using the Codex CLI, Gemini CLI, or both. A developer uses it for non-trivial changes, risky refactors, or security and performance-sensitive edits before opening or merging a PR. It detects the review scope, keeps the run read-only, presents findings by tool with highest severity first, then synthesizes where the models agree and disagree.1installs