Vague requirements lead to costly rework and missed deadlines. This skill transforms rough ideas into validated designs through Socratic dialogue, ensuring clarity before implementation begins.

An AI skill for Claude, Codex, and Claude Code

AI agents often claim success without actual verification. This skill prevents false completion claims by requiring fresh verification commands and evidence before any success assertion. It enforces discipline to run actual tests, show exit codes, and document proof.

Projects often lack documentation for directory structures. This skill automatically generates CLAUDE.md files with file responsibility tables when directories are created or modified, keeping documentation synchronized with code architecture.

Developers often guess at bug fixes, leading to regressions and wasted time. This skill enforces a rigorous 4-phase debugging process that requires root cause investigation before any code changes.

When working on long tasks with many tool calls, AI assistants often forget the original goal as it gets pushed out of the context window. This skill provides a protocol that forces reading of core documents at critical moments to keep the project goal front and center.

Teams struggle to remember and apply development workflow rules consistently. This skill provides quick access to all 10 Constitutional Articles, Phase -1 Gates, and common scenario answers from the CC-DevFlow Constitution v2.0.0.

Frontend projects often lack consistent patterns leading to maintenance issues and bugs. This skill provides comprehensive guidelines for React and TypeScript applications including modern patterns, file organization, and best practices.

Manual error tracking leaves blind spots in production. This skill enforces comprehensive Sentry v8 error capture across all services with standardized patterns for controllers, workflows, cron jobs, and database operations.

Deploy and manage Cloudflare serverless infrastructure directly from your AI assistant. This skill enables deployment of Workers, Pages, KV storage, R2 object storage, and D1 databases using the official wrangler CLI.

Managing terminal panes, tabs, and workspaces manually is repetitive and error-prone. This skill provides ready-to-use WezTerm CLI commands for automation and efficient terminal workflow management.

Agentic coding fails without detailed planning. This skill provides a proven methodology for creating thorough markdown plans through iterative multi-model refinement and conversion to executable beads.

Stop struggling to get screenshots to your remote AI sessions. giil downloads full-resolution images from iCloud, Dropbox, Google Photos, and Drive share links with one simple command.

Managing terminal windows and tabs manually slows down workflow. This skill provides a complete reference for Ghostty terminal emulator CLI commands. It enables quick window creation, tab management, split navigation, and configuration control without memorizing syntax.

Saving AI chat transcripts often breaks code blocks and creates messy filenames. This tool converts share links from ChatGPT, Gemini, Grok, and Claude into clean Markdown and HTML with preserved formatting and syntax highlighting.

Manual browser tasks consume too much time. This skill enables Claude to control your Chrome browser, automating repetitive form filling, data extraction, and workflow tasks using your existing login sessions.

Developers lose track of solutions found by AI coding assistants. This tool indexes sessions from Claude Code, Codex, Gemini, Cursor, and 8+ other agents into a unified searchable database.

Beads projects accumulate dependency graphs that are hard to navigate manually. BV computes PageRank, betweenness, cycles, and critical path metrics to help agents prioritize work efficiently.

Planning complex projects for AI agents is difficult because traditional task formats lack dependency awareness and self-contained context. This skill provides proven prompts to transform any markdown plan into detailed beads with explicit dependencies so agent swarms can execute work correctly without confusion.

Managing multiple AI agents in parallel creates coordination complexity. This skill provides exact prompts and workflows for coordinating agent swarms through NTM, Agent Mail, Beads, and BV tools.