run-parallel-agents-feature-debug

A UX/UI design specification skill that produces implementation-ready specs including user flows, component briefs, design tokens, accessibility constraints, and handoff guidance for frontend engineering agents.

An AI agent skill that estimates the development cost of a codebase, branch diff, or single commit by combining repo metrics, category-based productivity rates, organizational overhead, and optional Claude ROI analysis using helper scripts for LOC counting, git session analysis, and cost calculation.

A fast browser and native app automation skill for AI coding agents, providing persistent headless Chromium for web tasks and accessibility-API-based automation for Android, iOS, and macOS apps — supporting URL navigation, element interaction, form filling, JavaScript execution, and screenshots at ~100ms per command.

Filament v5 admin panel reference skill covering resources, forms, tables, actions, filters, relation managers, widgets, panels, and testing patterns with v5 namespace and API conventions.

A safe git merge workflow skill that handles branch merges, PR merges, conflict resolution, validation, rollback, and branch cleanup, with built-in readiness checks and worktree isolation for risky merges.

Scans the codebase and updates project memory files such as CLAUDE.md and imported reference docs with real project-specific routes, exports, flows, and implementation guidance.

A bug-finding skill that reviews branch diffs for real bugs and security issues, verifies each finding against surrounding code context, and produces a prioritized report of confirmed defects and vulnerabilities.

A skill that simplifies existing code without changing behavior by flattening nesting, extracting readable helpers, removing dead code, consolidating duplication, and improving naming, with built-in baseline verification to ensure correctness.

A skill that reviews implementation plans before execution, verifying markdown/JSON consistency, challenging scope and architecture, checking file paths against the codebase, auditing risk and test gaps, and delivering APPROVE/REVISE/REJECT verdicts. Runs as a forked review workflow with separate reasoning budget.

A structured bugfix skill that enforces a reproduce-diagnose-fix-verify loop with framework-specific references and category-based playbooks, supporting 15+ frameworks (React, Next.js, Express, Go, Rust, Laravel, etc.) to fix confirmed defects with minimal, root-cause-targeted changes.