GLM Tools, Skills & Agents

Comprehensive collection of AI platform skills, expert agents, system prompts, and development tooling from MiniMax, Super Z (GLM), z.ai, and the open-source community.

---

Quick Stats

Category	Count
Original Skills	4
External Skills	44
Community Skills	32
Agents	31
System Prompts	91
Commands	23
Hooks	23
MCP Integrations	9
Codebases	1
Total Files	500+

---

Repository Structure

GLM-Tools-Skills-Agents/
├── skills/                              # All skills
│   ├── minimax-experts/                 # 40 AI experts from MiniMax
│   ├── glm-skills/                      # Super Z/GLM multimodal skills
│   ├── zai-tooling-reference/           # Next.js 16 patterns
│   ├── ai-platforms-consolidated/       # Cross-platform reference
│   ├── external/                        # 44 external skills (superpowers, etc.)
│   └── community/                       # 32 community skills
│       ├── jat/                         # JAT task management (3 skills)
│       ├── pi-mono/                     # Pi coding agent (3 skills)
│       ├── picoclaw/                    # Go-based AI assistant (5 skills)
│       ├── dyad/                        # Local AI app builder (18 skills)
│       └── dexter/                      # Financial research CLI (1 skill)
├── agents/                              # Autonomous agents
│   ├── claude-codex-settings/           # 8 agents
│   └── community/                       # 19 community agents
│       ├── pi-mono/                     # 4 subagents (scout, planner, reviewer, worker)
│       └── toad/                        # 15 agent configs
├── system-prompts/                      # Leaked system prompts
│   ├── anthropic/                       # 15 Claude prompts
│   ├── openai/                          # 49 GPT prompts
│   ├── google/                          # 13 Gemini prompts
│   ├── xai/                             # 5 Grok prompts
│   └── other/                           # 9 misc prompts
├── commands/                            # Slash commands (23)
├── hooks/                               # Hook scripts (23)
│   ├── claude-codex-settings/           # 14 scripts
│   └── community/jat/                   # 9 scripts
├── prompts/                             # Prompt templates
│   └── community/pi-mono/               # 6 templates
├── mcp-configs/                         # MCP server configs
├── codebases/z-ai-tooling/              # Full Next.js 16 project
├── original-docs/                       # Source documentation
└── registries/                          # Skill registries

---

Skills Catalog

Original Skills (4)

Skill	Source	Description
minimax-experts	MiniMax	40 AI experts (Content, Finance, Dev, Career, Business, Marketing)
glm-skills	Super Z/GLM	Multimodal capabilities (ASR, TTS, VLM, Image/Video, PDF/DOCX/XLSX)
zai-tooling-reference	z.ai	Next.js 16, React 19, shadcn/ui, Prisma patterns
ai-platforms-consolidated	Various	Cross-platform comparison reference

External Skills (44)

From obra/superpowers, ui-ux-pro-max, claude-codex-settings:

• Development workflow: brainstorming, writing-plans, test-driven-development, systematic-debugging
• Git workflow: commit-workflow, pr-workflow, using-git-worktrees
• Quality assurance: verification-before-completion, requesting-code-review
• Tool integrations: azure-usage, gcloud-usage, supabase-usage, mongodb-usage, tavily-usage
• Plugin development: agent-development, command-development, hook-development, mcp-integration

Community Skills (32)

Source	Skills	Focus
jat	3	Task management (jat-start, jat-verify, jat-complete)
pi-mono	3	Coding agents (codex-cli, codex-5.3-prompting, interactive-shell)
picoclaw	5	Go assistant (github, weather, tmux, summarize, skill-creator)
dyad	18	Local app builder (swarm-to-plan, multi-pr-review, fix-issue, lint, etc.)
dexter	1	Financial research (dcf valuation)

---

Agents Catalog (31)

From claude-codex-settings (8)

• commit-creator, pr-creator, pr-reviewer (GitHub workflow)
• code-simplifier (pattern consistency)
• responsive-tester (viewport testing)
• agent-creator, plugin-validator, skill-reviewer (plugin dev)

From Community (23)

• pi-mono subagents (4): scout, planner, reviewer, worker
• toad agent configs (19): Claude, Codex, Gemini, Copilot, OpenCode, Goose, Kimi, etc.

---

System Prompts Catalog (91)

Anthropic (15)

• claude-opus-4.6.md, claude-code.md, claude-cowork.md
• claude-in-chrome.md, claude-for-excel.md
• Historical: claude-3.7-sonnet, claude-4.5-sonnet, claude-opus-4.5

OpenAI (49)

• GPT-5 series: gpt-5-thinking, gpt-5.1 (default/professional/nerdy/friendly/etc.), gpt-5.2-thinking
• Reasoning models: o3, o4-mini, o4-mini-high
• Tools: tool-python, tool-web-search, tool-deep-research, tool-canvas, tool-memory

Google (13)

• gemini-2.5-pro, gemini-2.5-flash, gemini-3-pro, gemini-3-flash
• gemini-workspace, gemini-cli, gemini-diffusion

xAI (5)

• grok-4, grok-4.1-beta, grok-3, grok-personas

Other (9)

• Notion AI, Raycast AI, Warp 2.0 Agent, Kagi Assistant, Sesame AI Maya

---

Hooks Catalog (23)

From claude-codex-settings (14)

• Code formatting: format_python_docstrings.py, prettier_formatting.py, markdown_formatting.py
• Git workflow: git_commit_confirm.py, gh_pr_create_confirm.py
• Web tools: websearch_to_tavily_search.py, webfetch_to_tavily_extract.py
• Notifications: notify.sh

From JAT (9)

• Session management: session-start-agent-identity.sh, pre-compact-save-agent.sh
• Signal tracking: post-bash-jat-signal.sh, user-prompt-signal.sh
• Activity logging: log-tool-activity.sh

---

Prompts/Templates (6)

From pi-mono:

• pr.md - Review PRs with structured analysis
• is.md - Analyze GitHub issues
• cl.md - Audit changelog entries
• codex-review-plan.md - Launch Codex to review plans
• codex-implement-plan.md - Codex implementation workflow
• codex-review-impl.md - Codex code review workflow

---

MCP Integrations (9)

Server	Package	Features
Azure	@azure/mcp	40+ Azure services
Google Cloud	gcloud-observability-mcp	Logs, metrics, traces
GitHub	Built-in	PRs, issues, workflows
Linear	linear-mcp	Issue tracking
MongoDB	mongodb-mcp-server	Database exploration
Playwright	@playwright/mcp	Browser automation
Slack	slack-mcp-server	Message search
Supabase	HTTP endpoint	Database, Auth, RLS
Tavily	tavily-mcp	Web search & extraction

---

Sources

Source	Type	Content
MiniMax Experts	Platform	40 AI experts
Super Z/GLM	Platform	Multimodal skills
obra/superpowers	GitHub	14 workflow skills
ui-ux-pro-max-skill	GitHub	1 design skill
claude-codex-settings	GitHub	29 skills, 8 agents, 23 commands
jat	GitHub	3 skills, 9 hooks
pi-mono	GitHub	3 skills, 4 agents, 6 prompts
picoclaw	GitHub	5 skills
dyad	GitHub	18 skills
dexter	GitHub	1 skill
toad	GitHub	19 agent configs
system_prompts_leaks	GitHub	91 system prompts
awesome-openclaw-skills	GitHub	3,002 skills registry
OS-Copilot	GitHub	Self-improvement framework
Prometheus	GitHub	Multi-agent bug fixing
zed	GitHub	Editor AI with MCP
skills.sh	Web	40+ skills catalog
buildwithclaude.com	Web	117 subagents

---

Installation

For Claude Code

cp -r skills/* ~/.claude/skills/

From skills.sh

npx skills add <skill-name>

---

Update History

Date	Changes
2026-02-13	Initial repository with GLM skills
2026-02-13	Added obra/superpowers (14 skills)
2026-02-13	Added ui-ux-pro-max-skill, claude-codex-settings
2026-02-13	Added community repos: jat, pi-mono, picoclaw, dyad, dexter
2026-02-13	Added 91 leaked system prompts (Anthropic, OpenAI, Google, xAI)
2026-02-13	Added toad agent configs, JAT hooks, pi-mono prompts

---

Updated: 2026-02-13