Downloads · 30 days
0
ejarbe/claude-skills
claude-skills is a machine learning model from ejarbe. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
1,253 skills for Claude Code — the largest personal collection of automation, development, reasoning, security, infrastructure, and creative skills for the Claude CLI agent.
Downloads · 30 days
0
Access
Public
Updated Feb 20, 2026
Repo size
109 MB
Likes
0
Public
Click a slice to open those files.
.gif88.8 MB · 47%
From the Hugging Face model README
1,253 skills for Claude Code — the largest personal collection of automation, development, reasoning, security, infrastructure, and creative skills for the Claude CLI agent.
skills/
├── 910 SaaS & API automations
├── 97 Azure SDK & cloud skills
├── 27 routers & orchestrators
├── 24 Google service integrations
├── 16 agent & multi-agent frameworks
├── 14 Zoho suite integrations
├── 11 reasoning & cognitive tools
├── 8 Hugging Face ML skills
├── 8 Obsidian knowledge management
├── 8 security & vulnerability scanning
├── 7 file format processors
├── 5 deployment targets
└── 100+ development, architecture, research & creative skills
Skills are markdown-based prompt modules that extend Claude Code's capabilities. Drop a skill folder into ~/.claude/skills/ and Claude gains specialized knowledge — from automating Slack workflows to running Semgrep scans to generating podcast audio.
Each skill directory contains:
| File | Purpose |
|---|---|
SKILL.md | Core instructions, workflows, and trigger conditions |
CLAUDE.md | Optional project-level overrides for Claude Code |
references/ | Supporting docs, examples, API schemas |
scripts/ | Executable helpers (Python, Node, shell) |
End-to-end integrations via Composio/Rube MCP — each skill knows the API surface, handles auth flows, and provides ready-to-use workflows.
<details> <summary>Highlights (click to expand)</summary>| Skill | What It Does |
|---|---|
slack-automation | Messaging, channels, search, reactions, threads |
notion-automation | Databases, pages, blocks, search, permissions |
github-automation | Issues, PRs, repos, actions, webhooks |
salesforce-automation | CRM records, SOQL, bulk operations |
hubspot-automation | Contacts, deals, pipelines, email sequences |
stripe-automation | Payments, subscriptions, invoices, refunds |
shopify-automation | Products, orders, inventory, fulfillment |
airtable-automation | Bases, tables, records, views, formulas |
jira-automation | Issues, sprints, boards, JQL queries |
discord-automation | Messages, channels, roles, server management |
twilio-automation | SMS, voice, WhatsApp via Telnyx/Twilio |
mailchimp-automation | Campaigns, audiences, templates, analytics |
zoho-crm-automation | Leads, contacts, deals, custom modules |
...and 897 more covering every major SaaS platform.
</details>Deep Azure integration skills across languages (Python, Java, .NET, Rust, TypeScript):
Advanced thinking frameworks that enhance Claude's problem-solving:
| Skill | Approach |
|---|---|
think | Non-linear reasoning + mental models + literate programming |
reason | Structured step-by-step logical analysis |
dialectical | Thesis-antithesis-synthesis reasoning |
critique | Systematic evaluation and counterargument generation |
hierarchical-reasoning | Multi-level decomposition for complex problems |
thought-based-reasoning | Chain-of-thought with explicit reasoning traces |
constraints | Constraint satisfaction for bounded optimization |
textbook-grounding | Academic source-based verification |
grounding-router | Routes to the right grounding strategy |
Meta-skills that route to the right specialist based on task context:
meta-router → Master router across all categories
├── agents-router → Agent framework selection
├── analysis-router → Data/code analysis dispatch
├── build-router → Build system selection
├── cli-router → CLI tool selection
├── commands-router → Command pattern matching
├── context-router → Context strategy selection
├── data-router → Data pipeline routing
├── development-router → Dev workflow dispatch
├── documentation-router → Doc format/tool selection
├── infrastructure-router → IaC/cloud routing
├── reasoning-router → Thinking strategy selection
├── research-router → Research method dispatch
├── skills-router → Skill discovery & suggestion
├── think-router → Cognitive approach selection
└── tools-router → Tool selection optimization
| Skill | Capability |
|---|---|
vulnerability-scanning/semgrep | SAST with custom rules and autofix |
vulnerability-scanning/codeql | Deep semantic code analysis |
vulnerability-scanning/variant-analysis | Finding vulnerability variants across codebases |
vulnerability-scanning/insecure-defaults | Detecting insecure default configurations |
security-threat-model | STRIDE-based threat modeling from code |
security-best-practices | OWASP Top 10, secure coding patterns |
security-ownership-map | Asset ownership and responsibility mapping |
osgrep | OS-level security grep patterns |
| Skill | Framework |
|---|---|
multi-agent-patterns | Architecture patterns for multi-agent systems |
multi-agent-coordination | Inter-agent communication protocols |
agents-v2-py | Python Agent SDK v2 |
agent-framework-azure-ai-py | Azure AI Agent Service |
agent-evaluation | Agent quality assessment methodologies |
agent-prompt-evolution | Iterative prompt improvement for agents |
agent-observability | Monitoring and debugging agent systems |
maker-framework | MAKER — consensus-voting multi-agent reasoning |
m365-agents-dotnet / m365-agents-py | Microsoft 365 Agents SDK |
hosted-agents-v2-py | Cloud-hosted agent deployment |
copilot-sdk | Microsoft Copilot SDK integration |
| Skill | Purpose |
|---|---|
software-architecture | System design and architecture guidance |
architecture-decision-records | ADR creation (Michael Nygard template) |
code-refactor / refactor | Structured refactoring workflows |
test-generation | Test infrastructure setup (Vitest, xUnit, pytest) |
test-driven-development | TDD red-green-refactor cycle |
property-based-testing | PBT patterns (Trail of Bits) |
testing-strategy | Test pyramid and coverage planning |
ci-cd-optimization | Pipeline performance tuning |
dependency-health | Dependency audit and update strategies |
technical-debt-management | Tech debt tracking and prioritization |
baseline-quality-assessment | Codebase health scoring |
database-optimizer | SQL optimization (PostgreSQL, MySQL, SQLite, SQL Server) |
api-design | REST/GraphQL API design patterns |
| Skill | Scope |
|---|---|
iac-terraform | Terraform/OpenTofu modules, testing, CI/CD |
k8s-troubleshooter | Kubernetes incident response & diagnostics |
observability-instrumentation | Logs, metrics, traces (OpenTelemetry, Prometheus, slog) |
cloudflare-deploy | Cloudflare Workers/Pages deployment |
vercel-deploy | Vercel project deployment |
netlify-deploy | Netlify site deployment |
render-deploy | Render service deployment |
| Skill | Capability |
|---|---|
hugging-face-model-trainer | Fine-tuning models on HF |
hugging-face-evaluation | Model benchmarking and eval |
hugging-face-datasets | Dataset creation and management |
hugging-face-tool-builder | Building HF Spaces/tools |
hugging-face-trackio | Experiment tracking |
dspy / dspy-code | Stanford's DSPy framework for LM programming |
prompt-engineering | Production prompt design and optimization |
context-engineering | Context window optimization for agents |
gpt-researcher | Autonomous web research agent |
fabric | 242+ Fabric CLI patterns for content processing |
| Skill | Function |
|---|---|
obsidian | Core Obsidian vault workflows |
obsidian-markdown | Advanced markdown formatting |
obsidian-batch | Bulk operations across vaults |
obsidian-data-importer | External data → Obsidian notes |
obsidian-developer | Plugin and theme development |
obsidian-devtools | Developer tooling for Obsidian |
obsidian-bases | Database-like views in Obsidian |
obsidian-process | Process documentation in vaults |
| Skill | Output |
|---|---|
suno-song | Suno V5 song generation prompts |
podcast-generation | AI podcast audio via Azure OpenAI Realtime |
imagegen | Image generation/editing via OpenAI API |
video-downloader | YouTube/platform video downloads |
youtube-transcript | Video transcript extraction |
canvas-design | Visual design on canvas |
algorithmic-art | Generative art patterns |
speech | Text-to-speech workflows |
transcribe | Audio-to-text transcription |
Handle documents natively within Claude Code:
pdf · docx · pptx · xlsx · csv-analysis · spreadsheet · json-canvas
| Skill | Use Case |
|---|---|
notion-automation | Core Notion API operations |
notion-knowledge-capture | Automated knowledge base building |
notion-meeting-intelligence | Meeting notes → structured Notion pages |
notion-research-documentation | Research → organized Notion docs |
notion-spec-to-implementation | Notion specs → code implementation |
Full coverage: gmail · google-drive · google-calendar · google-maps · google-search-console · google-classroom · google-admin · google-docs · google-sheets · google-slides · google-tasks · google-meet · google-photos · google-bigquery · google-analytics · google-ads · google-cloud-vision · and more.
# Copy one skill into your Claude Code skills directory
cp -r skills/slack-automation ~/.claude/skills/
# Clone and symlink the entire collection
git clone https://github.com/jarbitechture/claude-skills.git ~/claude-skills
ln -s ~/claude-skills/skills/* ~/.claude/skills/
# Example: install all security skills
for skill in vulnerability-scanning security-threat-model security-best-practices security-ownership-map osgrep; do
cp -r skills/$skill ~/.claude/skills/
done
User: "Send a message to #general in Slack saying deployment is done"
Claude Code:
1. Matches intent → loads slack-automation skill
2. Skill provides: API surface, auth flow, tool schemas
3. Claude executes: RUBE_SEARCH_TOOLS → find send_message → call it
4. Result: Message sent to #general ✓
Skills trigger automatically based on context, or can be invoked explicitly via /skill-name in Claude Code.
| Metric | Count |
|---|---|
| Total skills | 1,253 |
| Total files | 5,053 |
| Skills with reference docs | 187 |
| Skills with CLAUDE.md | 31 |
| Lines of instruction | 1.16M |
Skills sourced, adapted, and extended from open-source contributors:
| Source | Skills |
|---|---|
| antonbabenko/terraform-skill | Terraform/OpenTofu |
| trailofbits/skills | Security scanning, PBT |
| ahmedasmar/devops-claude-skills | DevOps, K8s |
| andronics/claude-plugin-adr | Architecture Decision Records |
| levnikolaevich/claude-code-skills | Development tools |
| QuestForTech-Investments/claude-code-skills | Automation skills |
| Composio | SaaS integration toolkits |
| Community & custom | Reasoning, orchestration, creative |
This is a personal collection. Individual skills may carry their own licenses from upstream sources. Check each skill's directory for attribution details.