Skip to content
/
Tier B A4 Markdown scaffold

CCPlugins (brennercruvinel)

ccplugins-cruvinel · brennercruvinel/CCPlugins · ★ 2.7k · last commit 2025-10-07

24 professional Claude Code commands that prevent AI overengineering and automate repetitive development tasks.

Best whenAI should never add attribution watermarks or signatures to professional codebases.
Skip ifCo-authored-by Claude in commits, Generated with Claude Code in PR descriptions
vs seeds
spec-kit(command-pack, structured workflows) but CCPlugins explicitly bans AI attribution watermarks (opposite of oddball's poli…
Primitive shape 24 total
Commands 24
00

Summary

CCPlugins (brennercruvinel)

A curated set of 24 professional Claude Code CLI slash commands (v2.5.2) designed to "automate the boring stuff" and save 2-3 hours per week on repetitive development tasks. CCPlugins is notable for explicit attribution policy (no "Co-authored-by Claude" watermarks), a multi-agent code review command that spawns Security/Performance/Quality/Architecture sub-agents, and conversational design philosophy where commands guide through structured phases rather than auto-executing. V2 is in active development due to Anthropic SDK updates making several v1 commands redundant. Currently maintained and tested against Opus 4 and Sonnet 4, with documented Kimi K2 compatibility.

differs_from_seeds: Most similar to spec-kit (clean command-pack format, structured workflow phases, no MCP or skill layer), but CCPlugins explicitly prohibits AI attribution watermarks (which spec-kit and openspec allow), includes a multi-agent review command (which no seed does within a command), and positions v2 as a deliberate architectural redesign rather than feature addition.

01

Overview

Overview — CCPlugins (brennercruvinel)

Origin

Created by Brenner Cruvinel. v2.5.2. Active development (last commit 2025-10-07). 2,705 stars, 157 forks. Python-based install scripts.

Philosophy (verbatim from README)

"Professional commands for Claude Code CLI that save 2-3 hours per week on repetitive development tasks."

"Built by a dev tired of typing 'please act like a senior engineer' in every conversation."

Problem statement (verbatim)

"😊 Ask Claude to fix a bug → Get 15 test files 😤 Request a simple refactor → Receive a dissertation on clean code 🤪 'Please add a button' → Complete UI framework rewrite 😭 Every conversation → 'Act like a senior engineer who doesn't overengineer'"

Conversational design principle

Commands use "conversational design" — structured phases that guide through analysis, recommendation, and validation rather than auto-executing. Each command is designed for predictable, incremental outcomes.

Anti-watermark policy (verbatim from /review command)

"Important: I will NEVER:

  • Add 'Co-authored-by' or any Claude signatures to commits
  • Add 'Created by Claude' or any AI attribution to issues
  • Include 'Generated with Claude Code' in any output
  • Modify git config or repository settings
  • Add any AI/assistant signatures or watermarks
  • Use emojis in commits, PRs, issues, or git-related content"

v2 development notice

V2 is in active development, treating several v1 commands as redundant due to Anthropic SDK improvements (native subagents, context memory, enhanced CLI). V2 will focus on what Claude Code still cannot do efficiently.

02

Architecture

Architecture — CCPlugins (brennercruvinel)

Distribution

Standalone GitHub repo. Install via shell script or Python install script.

Install

# Mac/Linux
curl -sSL https://raw.githubusercontent.com/brennercruvinel/CCPlugins/main/install.sh | bash

# Windows/Cross-platform
python install.py

# Manual
git clone https://github.com/brennercruvinel/CCPlugins.git
cd CCPlugins && python install.py

# Uninstall
./uninstall.sh
python uninstall.py

Required runtime

  • Python 3 (for install.py, uninstall.py)
  • Bash (for install.sh, uninstall.sh)
  • Claude Code CLI

Directory tree

.
├── commands/
│   ├── cleanproject.md
│   ├── commit.md
│   ├── contributing.md
│   ├── create-todos.md
│   ├── docs.md
│   ├── explain-like-senior.md
│   ├── find-todos.md
│   ├── fix-imports.md
│   ├── fix-todos.md
│   ├── format.md
│   ├── implement.md
│   ├── make-it-pretty.md
│   ├── predict-issues.md
│   ├── refactor.md
│   ├── remove-comments.md
│   ├── review.md
│   ├── scaffold.md
│   ├── security-scan.md
│   ├── session-end.md
│   ├── session-start.md
│   ├── test.md
│   ├── todos-to-issues.md
│   ├── understand.md
│   └── undo.md
├── install.py
├── install.sh
├── uninstall.py
└── uninstall.sh

Target AI tools

Claude Code (Opus 4, Sonnet 4). Documented as also working with Kimi K2.

Config files

None (commands installed directly to ~/.claude/commands/).

03

Components

Components — CCPlugins (brennercruvinel)

Commands (24 total)

Development Workflow

Command Purpose
/cleanproject Remove debug artifacts with git safety
/commit Smart conventional commits with analysis + pre-commit quality checks
/format Auto-detect and apply project formatter
/scaffold feature-name Generate complete features from patterns
/test Run tests with intelligent failure analysis
/implement url/path/feature Import and adapt code from any source with validation phase
/refactor Intelligent code restructuring with validation & de-para mapping

Code Quality & Security

Command Purpose
/review Multi-agent analysis (Security, Performance, Quality, Architecture sub-agents)
/security-scan Vulnerability analysis with extended thinking & remediation tracking
/predict-issues Proactive problem detection with timeline estimates
/remove-comments Clean obvious comments, preserve valuable docs
/fix-imports Repair broken imports after refactoring
/find-todos Locate and organize development tasks
/create-todos Add contextual TODO comments based on analysis results
/fix-todos Intelligently implement TODO fixes with context

Documentation

Command Purpose
/docs Generate/update project documentation
/explain-like-senior Explain code as a senior engineer would
/contributing Generate CONTRIBUTING.md

Project Management

Command Purpose
/understand Comprehensive project architecture analysis
/session-start Document session context in CLAUDE.md memory
/session-end Save session summary and progress
/make-it-pretty Visual/UX improvements
/todos-to-issues Convert TODOs to GitHub Issues
/undo Safe rollback with git integration

Skills

None.

Subagents

None (the /review command spawns conceptual "sub-agents" inline as roles, not formal subagent primitives).

Hooks

None.

Scripts

  • install.py — cross-platform install
  • install.sh — Mac/Linux install
  • uninstall.py — cross-platform uninstall
  • uninstall.sh — Mac/Linux uninstall

MCP servers

None.

05

Prompts

Prompts — CCPlugins (brennercruvinel)

Verbatim excerpt 1 — commands/commit.md (opening)

# Smart Git Commit

I'll analyze your changes and create a meaningful commit message.

**Pre-Commit Quality Checks:**
Before committing, I'll verify:
- Build passes (if build command exists)
- Tests pass (if test command exists)
- Linter passes (if lint command exists)
- No obvious errors in changed files

First, let me check if this is a git repository and what's changed:

```bash
# Verify we're in a git repository
if ! git rev-parse --git-dir > /dev/null 2>&1; then
    echo "Error: Not a git repository"
    echo "This command requires git version control"
    exit 1
fi
...

Prompting technique: Declarative pre-conditions + embedded bash verification blocks. The command declares what it will do before doing it, then includes specific bash scripts inline. This "narrate then execute" pattern creates a structured, auditable workflow.

Verbatim excerpt 2 — commands/review.md (anti-watermark policy + multi-agent)

# Code Review

I'll review your code for potential issues.

Let me create a checkpoint before detailed analysis:
```bash
git add -A  
git commit -m "Pre-review checkpoint" || echo "No changes to commit"

I'll use specialized sub-agents for comprehensive analysis:

  • Security sub-agent: Credential exposure, input validation, vulnerabilities
  • Performance sub-agent: Bottlenecks, memory issues, optimization opportunities
  • Quality sub-agent: Code complexity, maintainability, best practices
  • Architecture sub-agent: Layer separation, dependency direction, scalability patterns

...

Important: I will NEVER:

  • Add "Co-authored-by" or any Claude signatures to commits
  • Add "Created by Claude" or any AI attribution to issues
  • Include "Generated with Claude Code" in any output
  • Modify git config or repository settings
  • Add any AI/assistant signatures or watermarks
  • Use emojis in commits, PRs, issues, or git-related content

**Prompting technique**: Negative constraint injection (explicit "NEVER" list) + inline sub-agent role declaration. The NEVER list is a strong behavioral constraint using uppercase "NEVER" to signal non-negotiable rules. The sub-agent declarations are informal persona switches, not formal Claude Code subagent primitives.
09

Uniqueness

Uniqueness — CCPlugins (brennercruvinel)

differs_from_seeds

Most similar to spec-kit (command-pack format, structured phase-based workflow) and openspec (human-readable command descriptions). CCPlugins' most distinctive aspect vs seeds is the explicit no-watermark policy — refusing to add "Co-authored-by Claude" or "Generated with Claude Code" attribution. This is the opposite of the oddball team's attribution requirement. The multi-agent pattern in /review (four role-based inline sub-agents) is unusual in a pure command-pack. The v2 architectural redesign acknowledging Claude Code SDK evolution is also forward-looking among small command packs.

Positioning

"Professional commands for developers who are tired of AI overengineering." The 2,705 stars suggest this resonated — the problem statement ("ask for a button, get a framework rewrite") is a widely-felt pain. The anti-watermark stance is a deliberate differentiator for teams with attribution policies.

Observable failure modes

  • v2 in active development: several v1 commands may be redundant as of Claude Code's current SDK version.
  • No hooks: the pre-commit quality checks in /commit run because the command prompts Claude to run them, not via a PreToolUse hook — meaning they can be bypassed or skipped.
  • Inline sub-agents: the /review multi-agent pattern is a prompt trick, not a formal subagent delegation.

Cross-references

No explicit references to seeds or other frameworks.

04

Workflow

Workflow — CCPlugins (brennercruvinel)

Phases

Phase Description Artifact
Install python install.py ~/.claude/commands/*.md
Session start /session-start Session context in CLAUDE.md
Understand /understand Architecture analysis
Implement /scaffold, /implement, /fix-todos Code changes
Quality /test, /review, /security-scan Review output
Commit /commit Conventional commit
Session end /session-end Session summary in CLAUDE.md

Commit workflow (from /commit command)

  1. Verify git repository + detect changes
  2. Run pre-commit quality checks (build, tests, linter if configured)
  3. Analyze changes (file modifications, change type, scope)
  4. Generate conventional commit message
  5. Commit (staged files; stages modified files if nothing staged)

Review workflow (multi-agent approach)

The /review command instructs Claude to spawn four role-based "sub-agents":

  • Security sub-agent: Credential exposure, input validation, vulnerabilities
  • Performance sub-agent: Bottlenecks, memory issues
  • Quality sub-agent: Complexity, maintainability
  • Architecture sub-agent: Layer separation, dependency direction, scalability

These are inline persona-switching, not formal subagent primitives.

Approval gates

None automatic. The commit command creates a "pre-review checkpoint" commit before analysis, providing an implicit undo point.

06

Memory Context

Memory & Context — CCPlugins (brennercruvinel)

State storage

Session context stored in CLAUDE.md via /session-start and /session-end commands. No persistent external store.

Persistence

Project-level. CLAUDE.md survives sessions but is manually managed.

Compaction handling

Not addressed.

Context handoff

/session-start reads existing CLAUDE.md memory, /session-end writes session summary — creating a simple read/write session handoff via the project CLAUDE.md file.

07

Orchestration

Orchestration — CCPlugins (brennercruvinel)

Multi-agent

Pseudo-multi-agent. The /review command declares four sub-agent roles (Security, Performance, Quality, Architecture) but these are inline persona switches within a single Claude conversation, not formal Claude Code subagent primitives.

Orchestration pattern

Sequential (review phases run in order within one session).

Isolation mechanism

None.

Multi-model

No.

Execution mode

One-shot per command.

Consensus mechanism

None.

Prompt chaining

The /review command creates a git checkpoint commit first, then chains analysis → issue creation → optionally GitHub issues. The pre-review commit is an explicit chaining artifact.

08

Ui Cli Surface

UI/CLI Surface — CCPlugins (brennercruvinel)

Dedicated CLI binary

None. Install scripts are Bash/Python, not a named binary.

Local web dashboard

None.

IDE integration

Claude Code CLI. Documented as also working with Kimi K2.

Observability

None beyond CLAUDE.md session notes.

Install surface

curl -sSL https://raw.githubusercontent.com/brennercruvinel/CCPlugins/main/install.sh | bash

One-liner curl install.

Related frameworks

same archetype · same primary tool · same memory type

BMAD-METHOD ★ 48k
A4 Markdown scaffold

Provides a full agile delivery lifecycle with named expert-persona AI collaborators that elicit the human's best thinking rather…

Agent OS ★ 4.6k
A4 Markdown scaffold

Extracts implicit codebase conventions into token-efficient markdown standards files and injects them selectively into AI agent…

Claude Conductor ★ 367
A4 Markdown scaffold

Gives Claude Code a persistent, cross-linked, auto-analyzed documentation system so it retains codebase context across sessions.

Spec-Driver (Greenfield Spec-Driven Development) ★ 25
A4 Markdown scaffold

Prevents spec rot in AI-assisted development by making implementation changes flow back into evergreen, authoritative specs via…

Anthropic Knowledge Work Plugins ★ 16k
A4 Markdown scaffold

Role-specialized plugin bundles with live MCP connectors that turn Claude into a domain expert for enterprise knowledge workers.

Codex Integration for Claude Code (skill-codex) ★ 1.3k
A4 Markdown scaffold

Single Claude Code skill that handles Codex CLI invocation correctly (stdin blocking, thinking token suppression, session resume)…