Cycleaddict

Agents often skip documentation or use skills inconsistently under pressure. This skill provides a test-driven development framework for creating skills that agents actually follow. Build skills using RED-GREEN-REFACTOR cycles, then pressure-test them to ensure compliance even when agents face time pressure, sunk costs, or authority conflicts.

Developers struggle to break down complex features into clear, executable tasks. This skill generates comprehensive implementation plans with exact file paths, test code, and step-by-step instructions following TDD principles.

Prevents premature completion claims by requiring fresh verification evidence before commits or pull requests. This skill teaches Claude Code to always run verification commands and check output before making any success claims.

Claude sometimes skips skill checks and acts directly. This skill enforces disciplined skill invocation by requiring users to read relevant skill files before responding, ensuring proper workflows are followed.

Working on multiple features simultaneously requires constant branch switching, which disrupts context and slows progress. This skill creates isolated git worktrees that let you work on separate branches in parallel directories without switching, while automatically verifying gitignore safety and running clean test baselines.

Writing tests after code leads to bugs and poor design. This skill enforces strict TDD discipline with red-green-refactor cycles for bulletproof code.

Random fixes waste time and create new bugs. This skill provides a 4-phase framework to find root causes before attempting any solution. Stop guessing and start investigating.

Implement complex projects by dispatching subagents for independent tasks. Each task gets two-stage review ensuring spec compliance and code quality. Stay in the same session for fast iteration.

Manual code review is time-consuming and inconsistent. This skill dispatches a structured code-reviewer subagent to systematically evaluate code changes against requirements, architecture standards, and production readiness criteria.

Code review feedback requires technical verification, not blind implementation. This skill guides AI agents through evaluating feedback critically, asking clarifying questions, and pushing back when necessary with technical reasoning.

Traditional AI coding is destructive. Once you change direction, previous work is overwritten. Parametric Scribe creates a non-linear editing experience for code, recording each task as a replayable step that can be modified and intelligently reapplied to new contexts.

After implementing a feature, deciding how to integrate the work can be unclear. This skill provides a structured process to verify tests, present clear options for merge or PR, and safely execute the chosen workflow with confirmation prompts.

Implementation plans often fail when executed without structure. This skill provides a batch-based workflow with mandatory review checkpoints between work batches for safer, more collaborative development.

Multiple test failures waste time when investigated sequentially. Parallel agent dispatch sends independent investigations to multiple AI agents, solving several problems in the time it takes to solve one.

Starting implementation without clear requirements leads to rework and missed expectations. This skill guides structured brainstorming conversations to understand intent, explore alternatives, and validate designs before writing code.