ww-w-ai/bkit-claude-code

Audit shows recent audit log entries, decision traces, and session history for AI-assisted workflows. Use it to inspect actions, trace feature decisions, summarize activity, or search historical events.

BKIT provides a Claude Code command set for planning, development phases, review, deployment, and project setup. Use it to browse available functions with /bkit or bkit help.

Bkend Auth explains bkend.ai authentication flows for email, social, and magic-link sign-in, plus JWT refresh, sessions, and RBAC. Use it when adding sign-up, login, password reset, or access control code.

Bkend Cookbook provides project tutorials for bkend.ai apps from todo to SaaS. It also lists common implementation errors and fixes for authentication, uploads, permissions, and sync issues.

Bkend Data manages bkend.ai tables and records with CRUD operations, filtering, sorting, relations, and indexes. Use it when working with database schemas or data via MCP or the REST API.

Bkend Quickstart explains how to connect bkend.ai to Claude Code, understand its organization/project/environment model, and create a first project. Use it when setting up MCP access or orienting to the platform.

Bkend Storage explains how to upload, download, and manage bkend.ai files with presigned URLs, CDN links, visibility levels, and multipart uploads. Use it when implementing file storage flows or file management API calls.

Bkit Evals runs a skill's eval suite through evals/runner.js, validates the skill name, captures output, and saves structured results. Use it to list available evals or assess a skill's pass/fail output.

Bkit Explore scans installed bkit skills, agents, and evals from local files and shows them by category or level. Use it to find what is installed and locate the right skill quickly.

Bkit Rules defines automatic PDCA, level detection, agent triggering, and code quality rules for Claude Code work. Use it to guide feature work, bug fixes, task classification, and Sprint phase control.

Bkit Templates provides PDCA and sprint document templates for planning, design, analysis, and reporting. Use it when generating feature or sprint documentation in a structured workflow.

By-The-Way captures improvement ideas, bug patterns, and skill requests without interrupting the current task. Use it to record, list, analyze, promote, and review suggestions from Claude Code sessions.

CC Version Analysis researches Claude Code CLI changes between versions, maps them to bkit, and produces a Korean impact report. Use it when upgrading CC to assess risks, opportunities, and required follow-up analysis.

Claude Code Learning helps configure Claude Code settings, commands, hooks, and workflows. Use it to learn features, generate setup files, and review upgrades.

Code Review analyzes code for bugs, security issues, performance problems, and style concerns. Use it to review a file, directory, staged changes, or a pull request and get actionable feedback.

Control Panel shows bkit automation level, trust score, and active guardrails. Use it to view status, pause or resume automation, or change the trust-gated level from L0 to L4.

Deploy generates CI/CD pipeline files and guides feature promotion to dev, staging, or prod with level-based gates. Use it when you need environment-specific deployment flow, rollback, or promotion checks.

Desktop App Development covers building cross-platform desktop apps with web technologies using Electron or Tauri. Use it when creating Windows, macOS, and Linux apps with native file, menu, and window access.

Development Pipeline explains the 9-phase path from schema and conventions through mockups, API, UI, SEO/security, review, and deployment. Use it when starting a new project or when `/pipeline-*` commands ask for the order of work.

Dynamic supports full-stack app setup with bkend.ai for authentication, database access, and API integration. Use it to bootstrap and guide Next.js apps that need login, data, and backend service wiring.

Enterprise Architecture initializes and guides large-scale systems using microservices, Kubernetes, Terraform, and AI-native workflows. Use it for multi-feature initiatives that need monorepo structure, infra templates, and release planning.

Mobile App Development guides building iOS and Android apps with cross-platform frameworks like React Native, Expo, and Flutter. Use it when starting or structuring a mobile app from a web development background.

PDCA Batch manages multiple PDCA features as independent parallel cycles with status, planning, and queue handling. Use it when you need to track several features at once without sharing scope, budget, or timeline.

PDCA Fast Track auto-approves Checkpoints 1-8 when fast-track is enabled, Trust is at least 80, and a design doc exists. It falls back to L2 and manual review when any precondition fails.

PDCA Watch renders a read-only live dashboard for an active PDCA feature, updating every 30 seconds when loop support is available. It reads pdca-status.json and token-ledger.ndjson to show phase, match rate, iteration, tokens, and cost.