By zhengxuyu
Personality distillation, memory, and engineering workflow skills for AI coding agents
Add a new feature with design, TDD, and PR workflow
Audit a codebase for security vulnerabilities, code quality issues, and architecture concerns
Use when the user says "autopilot", "auto mode", "替我回答", or "自动回答". Enables autonomous decision-making by answering Claude's own questions based on the user's personality profile.
Write personal blog posts with the rhythm, structure, and voice of top-tier essayists (Paul Graham, Dan Luu, Henrik Karlsson, Ben Kuhn, etc.)
Fix a bug with systematic debugging, TDD, and PR workflow
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
My daily-driver AI coding agent skills. Works with Claude Code, Codex, OpenClaw, Hermes Agent, etc. Take what you find useful.
Solves the blank-slate problem — agents forget everything between sessions.
principle loads ~/.claude/principles.md at session start and silently extracts new signals from your corrections and decisions as you work. After a few sessions, a fresh agent doesn't need re-training. autopilot goes further — the agent consults your profile to answer its own questions instead of interrupting you, and logs every decision. memory auto-saves a working memory (what you're doing, decisions made, blockers) via session hooks and restores it on next start. Important stuff gets promoted to long-term memory that persists across projects.
| Skill | What it does |
|---|---|
principle | Extracts your style, decisions, quality bar into a reusable profile |
autopilot | Agent answers its own questions from your profile, doesn't interrupt |
memory | Working memory (daily notes) + long-term memory across sessions |
Guardrails for agents. Every skill enforces TDD, branch isolation, and PR gates.
| Skill | What it does |
|---|---|
add-feat | New feature — design, TDD, PR |
fix-bug | Root cause, regression test, PR |
hotfix | Emergency — smallest fix, fast PR |
minor-change | Quick tweak — branch, change, test, PR |
refactor | Restructure — tests green before and after |
| Skill | What it does |
|---|---|
pr-review | Isolated worktree review, structured report |
git-release | Tag, release notes, GitHub release |
update-deps | Changelogs, compatibility tests, PR |
| Skill | What it does |
|---|---|
init-project | Bootstrap repo, structure, tooling |
setup-ci | GitHub Actions via single-question flow |
watchdog | Monitor CI/CD, deployments, long-running jobs |
audit | Security, quality, architecture review |
Plugin install. Gets all skills + session hooks (personality distillation, memory auto-save/restore). superpowers will be auto-installed on first session if not already present.
claude plugin marketplace add zhengxuyu/zforge
claude plugin install zforge
Skills only (no hooks). Personality and memory skills still work but need manual invocation — no auto-inject at session start.
npx skills add zhengxuyu/zforge
Clone and symlink into your agent's skill directory.
git clone https://github.com/zhengxuyu/zforge.git ~/.zforge
ln -s ~/.zforge/skills ~/.agents/skills/zforge
MIT
npx claudepluginhub zhengxuyu/zforge --plugin zforgeCore skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Autonomous improvement engine for Claude Code. Runs an unbounded modify-verify-keep/discard loop against any mechanical metric. 10 subcommands: plan, debug, fix, security, ship, scenario, predict, learn, and reason.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Feature development with code-architect/explorer/reviewer agents, CLAUDE.md audit and session learnings, and Agent Skills creation with eval benchmarking from Anthropic.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review