II.
Page JSON
Structured · livepage:library-claudekit
ClaudeKit Methodology (Library) json
Inspect the normalized record payload exactly as the atlas UI reads it.
{
"id": "page:library-claudekit",
"_kind": "Page",
"_file": "wiki/library/claudekit.md",
"_cluster": "wiki",
"attributes": {
"nodeKind": "Page",
"title": "ClaudeKit Methodology (Library)",
"displayName": "ClaudeKit Methodology (Library)",
"slug": "library/claudekit",
"articlePath": "wiki/library/claudekit.md",
"article": "\n# ClaudeKit Methodology\n\nA coding safety net with real-time error prevention, checkpoint management, and specialized AI subagent orchestration. Adapted from [ClaudeKit](https://github.com/carlrannaberg/claudekit) by Carl Rannaberg.\n\n## Core Principles\n\n1. **Event-Driven Safety**: Hook system provides real-time error prevention across PreToolUse, PostToolUse, UserPromptSubmit, and Stop events\n2. **File Guard Protection**: 195+ patterns across 12 categories block sensitive file access with bash pipeline analysis\n3. **6-Agent Parallel Review**: Architecture, security, performance, testing, quality, and documentation reviewed concurrently\n4. **Spec-Driven Development**: Full specification lifecycle from codebase research to 6-phase iterative execution\n5. **Parallel Research**: 5-10 concurrent research agents for up to 90% faster results\n6. **Checkpoint Management**: Git-backed state management for safe rollback\n7. **Session Isolation**: All hook changes are session-scoped\n8. **Context Engineering**: Codebase mapping and thinking-level enhancement for improved reasoning\n\n## Process Files\n\n| File | Process ID | Description |\n|------|-----------|-------------|\n| `claudekit-orchestrator.js` | `methodologies/claudekit/claudekit-orchestrator` | Main entry: setup hooks, file-guard, codebase-map, dispatch commands, safety checks, checkpoints |\n| `claudekit-code-review.js` | `methodologies/claudekit/claudekit-code-review` | 6-agent parallel code review with weighted scoring and recommendations |\n| `claudekit-spec-workflow.js` | `methodologies/claudekit/claudekit-spec-workflow` | Spec creation from research + 6-phase execution (implement, test, review, improve, commit, track) |\n| `claudekit-research.js` | `methodologies/claudekit/claudekit-research` | Parallel research orchestration with 5-10 concurrent agents |\n| `claudekit-safety-pipeline.js` | `methodologies/claudekit/claudekit-safety-pipeline` | Hook system: file-guard, typecheck/lint/test checks, checkpoints, profiling |\n\n## Agents (8)\n\n| Agent | Role | Used In |\n|-------|------|---------|\n| `code-review-coordinator` | Multi-agent review aggregation | Code Review |\n| `security-analyst` | Security vulnerability detection | Code Review |\n| `performance-analyst` | Performance bottleneck analysis | Code Review |\n| `testing-specialist` | Test coverage and quality | Code Review |\n| `architecture-reviewer` | Architecture quality assessment | Code Review |\n| `research-coordinator` | Parallel research orchestration | Research |\n| `spec-architect` | Specification design and creation | Spec Workflow |\n| `triage-specialist` | Problem diagnosis and routing | Orchestrator |\n\n## Skills (8)\n\n| Skill | Purpose |\n|-------|---------|\n| `checkpoint-management` | Git-backed state management |\n| `file-guard` | 195+ pattern file access protection |\n| `code-review-orchestration` | 6-agent parallel review |\n| `spec-creation` | Feature specification from research |\n| `spec-execution` | 6-phase iterative implementation |\n| `research-orchestration` | Parallel multi-agent research |\n| `hook-management` | Session-scoped hook lifecycle |\n| `codebase-mapping` | Automatic project indexing |\n\n## Slash Commands (Mapped to Processes)\n\n| Command | Maps To |\n|---------|---------|\n| `/code-review [target]` | `claudekit-code-review` |\n| `/spec:create [feature]` | `claudekit-spec-workflow` (create mode) |\n| `/spec:execute [file]` | `claudekit-spec-workflow` (execute mode) |\n| `/research [query]` | `claudekit-research` |\n| `/checkpoint:create` | `claudekit-orchestrator` checkpoint task |\n| `/checkpoint:restore` | `claudekit-orchestrator` checkpoint task |\n| `/validate-and-fix` | `claudekit-safety-pipeline` |\n| `/hook:enable/disable/status` | `claudekit-orchestrator` hook management |\n\n## Hook System\n\n| Event | Hooks |\n|-------|-------|\n| PreToolUse | file-guard |\n| PostToolUse | typecheck-changed, lint-changed, test-changed, check-comment-replacement, check-unused-parameters |\n| UserPromptSubmit | codebase-map, thinking-level |\n| Stop | create-checkpoint, typecheck-project, lint-project, test-project, self-review |\n\n## Attribution\n\nAdapted from [ClaudeKit](https://github.com/carlrannaberg/claudekit) by Carl Rannaberg.\n",
"documents": [
"specialization:claudekit"
]
},
"outgoingEdges": [
{
"from": "page:library-claudekit",
"to": "specialization:claudekit",
"kind": "documents"
}
],
"incomingEdges": [
{
"from": "page:index",
"to": "page:library-claudekit",
"kind": "contains_page"
}
]
}