# coding-dev-skill

**Repository Path**: felixagents/coding-dev-skill

## Basic Information

- **Project Name**: coding-dev-skill
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-10-06
- **Last Updated**: 2026-10-08

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Coding Dev Skill

Collection of software development skills, providing complete software development lifecycle support for AI Agents.

## Project Introduction

Coding Dev Skill is an AI-driven software development skill library, personalizing and merging the core skills from obra/superpowers and mattpocock/skills into one coherent methodology engine. It covers the complete workflow from requirements analysis, design planning, coding implementation to test deployment. Each skill is an independent AI capability module that can be invoked by AI assistants to complete specific development tasks.

The library is organized around a **main flow (idea → production)**, supported by two on-ramps (issue triage, foggy-project planning) and a set of supporting skills underneath. Not sure where to start? Just ask **skill-router**.

## Skill Overview

### Main Flow: Idea → Production

| Skill | Description |
|------|------|
| **brainstorming** | Brainstorming & design planning: before any creative work, explore user intent, requirements, and design; expands into the PRD flow when needed |
| **prd-write** | PRD writing: analyze requirements, break down tasks, and produce executable high-quality Product Requirements Documents (docs/prds/) |
| **prd-review** | PRD review: rigorous, zero-tolerance-for-ambiguity expert review, catching missing scenarios, vague acceptance criteria, and technical risks |
| **grill-with-docs** | Solution stress testing (in-repo): interrogate plans, distill domain language into GLOSSARY.md, and record hard-to-reverse decisions as ADRs |
| **writing-plans** | Writing implementation plans: turn consensus into a TSP (docs/TSPs/) that records decisions, not transcripts of code |
| **executing-plans** | Executing plans: personally execute a TSP task by task with task-start/task-done loops, progress ledger, and the only four stop conditions |
| **subagent-driven-development** | Subagent-driven development: delegate each task to a subagent (brief → implement → review → ledger), orchestrating the build instead of driving it |
| **test-driven-development** | TDD: test-driven development, one red-green slice at a time, building features and fixing bugs test-first |
| **requesting-code-review** | Requesting code review: two-axis review of the diff (standards + spec), judged against what a reasonable user would expect |
| **verification-before-completion** | Verification before completion: run verification commands and confirm output before claiming done — evidence, not feelings |
| **finishing-a-development-branch** | Finishing a branch: after tests pass, present merge / PR / keep / discard options and execute the choice |
| **retro** | Retro: review the session and improve the agent's environment (navigation pointers, automated checks, coding standards), not the code |

### On-Ramps

| Skill | Description |
|------|------|
| **triage** | Issue triage: run issues and external PRs through a triage state machine, producing agent-ready briefs (AGENT-BRIEF) |
| **wayfinder** | Foggy-project planning: map a too-big-for-one-session effort as decision tickets on an issue tracker, producing decisions rather than deliverables |

### Supporting Skills

| Skill | Description |
|------|------|
| **using-superpowers** | Session auto-entry: pulls relevant skills into context at the start of work |
| **skill-router** | Skill router: describe your situation, and it tells you which path to take and which skills to invoke in what order |
| **grill-me** | Solution stress testing (standalone): sustained interrogation of a proposal when no repo is available |
| **receiving-code-review** | Receiving review feedback: verify technically before implementing; no performative agreement |
| **using-git-worktrees** | Git worktrees: create isolated, safe worktrees for feature work or bug fixes |
| **wait-what** | Message repair: re-articulate a poorly communicated message with the missing context |
| **handoff** | Context handoff: compress the current conversation into a handoff document for a fresh agent |

### Domain Skills

| Skill | Description |
|------|------|
| **git-commit** | Git commit: generate commit messages following the Conventional Commits spec |
| **project-overview** | Project overview: run it first when taking over an unfamiliar repo, to get an architecture and directory tour |
| **pr** | PR body writing: describe a change with a structured body of summary diagram, before/after evidence, and merge risk |
| **pstest** | API automation testing: write and run YAML test cases with the pstest CLI, import OpenAPI, generate reports and mock data |
| **pstest-ui** | UI automation testing: Playwright keyword-driven UI testing with the pstest ui CLI, with Excel/natural-language project generation |
| **frontend-design** | Frontend design: guidance for distinctive, intentional visual design when building or reshaping UI (aesthetic direction, typography, de-templating) |

## Quick Start

### Install Skills

A skill is a directory containing a `SKILL.md`; place it into your AI tool's skill directory (Claude Code as an example):

```bash
# Install a single skill at user level
cp -r skills/brainstorming ~/.claude/skills/

# Or install all skills
cp -r skills/* ~/.claude/skills/
```

Once installed, `using-superpowers` pulls relevant skills into context at session start; you can also invoke them directly via the Skill tool in conversation, or just say "help me route" to let `skill-router` point the way.

### API Testing with pstest

```bash
# Generate test cases
pstest generate --spec openapi.json

# Run tests
pstest run --config config.yaml

# Generate report
pstest report --output report.html
```

### UI Testing with pstest-ui

```bash
# Create a new test project
pstest ui new-project my-system --cookie "xxx" --url "https://..."

# Run tests
pstest ui run --project ./my-system

# Debug mode
pstest ui debug --project ./my-system
```

## Project Structure

```
.
├── skills/
│   ├── brainstorming/                    # Brainstorming & design
│   ├── prd-write/                        # PRD writing
│   ├── prd-review/                       # PRD review
│   ├── grill-with-docs/                  # Solution stress testing (in-repo)
│   ├── grill-me/                         # Solution stress testing (standalone)
│   ├── writing-plans/                    # Writing implementation plans (TSP)
│   ├── executing-plans/                  # Executing plans
│   ├── subagent-driven-development/      # Subagent-driven development
│   ├── test-driven-development/          # TDD
│   ├── requesting-code-review/           # Requesting code review
│   ├── verification-before-completion/   # Verification before completion
│   ├── finishing-a-development-branch/   # Finishing a development branch
│   ├── retro/                            # Retro
│   ├── triage/                           # Issue triage
│   ├── wayfinder/                        # Foggy-project planning
│   ├── using-superpowers/                # Session auto-entry
│   ├── skill-router/                     # Skill router
│   ├── receiving-code-review/            # Receiving review feedback
│   ├── using-git-worktrees/              # Git worktrees
│   ├── wait-what/                        # Message repair
│   ├── handoff/                          # Context handoff
│   ├── git-commit/                       # Git commit
│   ├── project-overview/                 # Project overview
│   ├── pr/                               # PR body writing
│   ├── pstest/                           # API testing
│   ├── pstest-ui/                        # UI testing
│   └── frontend-design/                  # Frontend visual design
```

## Documentation

- [Skill Router](./skills/skill-router/SKILL.md) — start here when unsure which skill to use
- [Phase Boundaries Decision Tree](./skills/skill-router/PHASE-BOUNDARIES.md) — the full decision tree for continue / clear / handoff / subagent / compress
- [pstest-ui User Guide](./skills/pstest-ui/USER_GUIDE.md)
- [pstest Skill Documentation](./skills/pstest/README.md)
- [Skill Authoring Template](./docs/SKILL-TEMPLATE.md) — single source of truth for writing SKILL.md

## License

MIT License