VividKit Crystal clear AI coding
VI
What is ClaudeKit? ClaudeKit CLI Guide Claude Mechanics Coexistence How ClaudeKit Works Inside ClaudeKit Commands Cheat Sheet Command Finder Workflow Recipes Accounts Switcher Mobile Coding Remote Control IDE Config Migrate to Other Tools Session Recovery UI/UX Guide Permission Modes Custom Hooks Fix from Logs Promotions

How ClaudeKit Works

ClaudeKit is not a standalone app or a new IDE. It is an advanced Framework integrated directly into your native Claude Code configuration (.claude) on your machine, transforming your terminal chat into a professional task automation system.

1. Hooks (Background System)

Background scripts that run automatically. They initialize configs and project rules so you don't have to copy-paste them manually.

2. Skills (Pre-built Knowledge)

Specialized knowledge packages (SKILL.md) with prompts, references and scripts. The primary execution unit when you type /ck:* commands.

3. Agents (Specialists)

Specialized models (Planner, Tester, Reviewer) with internal prompts and tools, focusing on a single task domain.

4. Workflows (Pipelines)

Pipelines of Hooks, Skills and Agents running sequentially from your command input to handle complex workflows end-to-end.

Let's see how Hooks, Skills, Agents and Workflows coordinate a real command. Pick a workflow below and press Play to explore the internal mechanism.

Guide coming soon...

VividKit is finalizing the detailed guide for Marketing Kit.

No combos for this kit

Guide coming soon...

VividKit is finalizing the detailed guide for Marketing Kit.

Step / Playing...

/ck:brainstorm

Solution brainstormer with trade-off analysis and brutal honesty. Design-first, no code.

HARD GATE: No code, no scaffolding, no implementation until design is presented and user approves. Applies to every session regardless of complexity.

PROCESS FLOW

1

Scout

Read project docs & code patterns

2

Discover

Ask clarifying questions via AskUser

3

Scope

Decompose if 3+ independent concerns

4

Research

Parallel researcher agents + web search

5

Analyze

Evaluate 2-3 approaches with pros/cons

6

Debate

Challenge assumptions, present options

7

Consensus

Align on chosen approach

8

Report

Write markdown summary

9

Next?

Ask: create plan? → /plan

Core Principles

  • › YAGNI — You Aren't Gonna Need It
  • › KISS — Keep It Simple, Stupid
  • › DRY — Don't Repeat Yourself
  • › Brutal honesty over diplomacy
  • › Challenge every assumption

Expertise Areas

  • › System architecture & scalability
  • › Risk assessment & mitigation
  • › UX/DX optimization
  • › Technical debt management
  • › Performance bottleneck identification

SKILL & AGENT STACK

Skills and agents used during sessions:

planner agent docs-manager agent WebSearch ck:scout ck:docs-seeker ck:ai-multimodal ck:sequential-thinking psql
Agent Skill Tool • All optional — activated based on context

OUTPUT

Markdown Summary Report

plans/reports/ brainstorm-YYMMDD-HHMM-slug.md
Problem statement Evaluated approaches Final recommendation Implementation risks Success metrics Next steps

/ck:plan

Implementation planner with research phases, red-team review, and task hydration. Creates actionable phase files.

PROCESS FLOW

1

Pre-Check

Check Plan Context + active plan

2

Mode

Auto-detect or explicit flag

3

Research

Spawn researcher agents

4

Analyze

Read docs, scout codebase

5

Plan

Write plan.md + phase files

6

Red Team

Adversarial review (2-4 reviewers)

7

Validate

Critical questions interview

8

Hydrate

Create Claude Tasks from phases

9

Handoff

Output /cook command

WORKFLOW MODES

Flag Mode Research Red Team Validation Cook Flag
--auto Auto-detect Follows mode Follows mode Follows mode Follows mode
--fast Fast Skip Skip Skip --auto
--hard Hard 2 researchers Yes Optional —
--deep Deep 3+ scouts/phase Yes Forced —
--parallel Parallel 2 researchers Yes Optional --parallel
--two Two approaches 2+ researchers After select After select —

--tdd (tests-first per phase) and --no-tasks (skip task hydration) can combine with any mode.

SPECIAL OPERATIONS

red-team

2-4 code-reviewer agents challenge plan assumptions, find blind spots, test edge cases. Run as /plan red-team.

validate

Critical questions interview (3-8 questions). Ensures plan completeness before implementation begins.

archive

Write journal entry documenting decisions. Archive completed/abandoned plans for future reference.

/ck:cook

Implementation engine that follows plans, runs tests, simplifies code, and reviews quality. The workhorse of ClaudeKit.

PROCESS FLOW

1

Intent

Detect mode from input

2

Research

Scout codebase (skip if fast)

3

Plan

Create plan.md + phases

4

Implement

Execute phase tasks

5

Test

Run tests via tester agent

6

Review

Code review via reviewer agent

7

Finalize

Sync plan + docs + commit

Core Principles

  • › Follow the plan exactly
  • › Test before review
  • › Delegate to subagents (tester, reviewer)
  • › Never skip quality checks
  • › Finalize with docs + commit

Expertise Areas

  • › Full-stack implementation
  • › Test-driven development
  • › Subagent orchestration
  • › Quality assurance
  • › Git workflow management

SKILL & AGENT STACK

Skills and agents used during sessions:

tester agent code-reviewer agent code-simplifier agent git-manager agent ck:scout ck:test ck:code-review
Agent Skill Tool • All optional — activated based on context

OUTPUT

Implementation Complete

Commits + updated plan status
Code changes committed Tests passing Quality reviewed Plan phases marked complete

/ck:fix

Structured debugging pipeline with root cause analysis, fix implementation, and regression prevention.

PROCESS FLOW

1

Scout

Gather evidence: logs, stack traces

2

Diagnose

Root cause analysis (no guessing)

3

Route

Assess complexity → workflow

4

Fix

Implement the fix

5

Verify

Run tests + add regression test

6

Finalize

Docs + commit + journal

Core Principles

  • › Find root cause, not symptoms
  • › One bug, one commit
  • › Always add regression test
  • › Document the fix
  • › Never ignore test failures

Expertise Areas

  • › Root cause analysis
  • › Debugging techniques
  • › Test coverage expansion
  • › Error handling patterns
  • › Log analysis

SKILL & AGENT STACK

Skills and agents used during sessions:

debugger agent tester agent ck:scout ck:test ck:sequential-thinking
Agent Skill Tool • All optional — activated based on context

OUTPUT

Bug Fix Report

Inline summary + commit
Root cause identified Fix implemented Tests passing Regression test added

Pipeline

/

Keyboard: Space = Play/Pause · ← → = Prev/Next step

Ready to experience it yourself?

Install ClaudeKit Now Explore All Commands