~200-Skill Operating System for AI-Native Enterprises
Agentic Enterprise OS is not a collection of prompts — it's a modular, layered operating system for AI-native enterprises. It provides:
- ~198 specialized skills (one
SKILL.mdper domain folder) across the original 7-layer model plus major expansions - 8 workflow chains for end-to-end operations (YAML templates in
.workflows/, skills.sh install hints) - 4 execution modes for different task complexities
- Cross-cutting validation via
build-validator - Persistent memory via
ai-memory-developer - Dynamic routing via
ai-skill-manager
Recent expansions add depth in actuarial and insurance, cloud and FinOps, SRE and resilience, predictive analytics, ML safeguards and agentic AI, AML and financial crime, supply chain, OT/ICS and networking, and a broader security portfolio (CTI, pentest, formal methods, vendor risk, executive security roles).
This package is installable from the Agent Skills directory using the open-source skills CLI (npx skills add — no global install required).
Source: github.com/daemon-blockint-tech/Agentic-Enteprises-Skill
Run from the root of the project where you want skills available (your app repo, not necessarily this catalog clone):
# Browse ~197 skills in this package
npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill --list
# Install specific skills for Cursor
npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill \
--skill cloud-architect \
--skill compliance-engineer \
--agent cursor -y
# Interactive picker
npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill --agent cursorAfter install, start a new Cursor session so agents pick up the installed SKILL.md files. See Skills for Cursor.
If the skills.sh package page shows fewer skills than this repo, use npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill --list for the full catalog—the directory index can lag behind GitHub.
The same CLI targets Claude Code, Codex, GitHub Copilot, Windsurf, Gemini, and others:
npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill --agent claude-code -y
npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill --agent '*' -y # all supported agentsMore: skills.sh documentation.
| Approach | When to use |
|---|---|
npx skills add … |
Install skills into another project's agent paths (recommended for day-to-day work in Cursor, etc.) |
| Clone this repo | Browse the catalog, contribute skills, or point an agent at <skill-name>/SKILL.md in-tree |
| Fork | Maintain a private or customized enterprise skill portfolio |
Leaderboard ranking uses anonymous install telemetry from the CLI (how skills are ranked). Opt out: DISABLE_TELEMETRY=1.
User: "Design a data warehouse schema for sales analytics"
↓
Route: data-warehouse-engineer
↓
Execute → Return
User: "Build end-to-end analytics for customer churn"
↓
Chain: data-architect → data-warehouse-engineer → analytics-data-engineer → data-scientist → bi-analyst
↓
Execute sequentially → Return
User: "Audit our security and optimize our data pipeline"
↓
Parallel: offensive-security-analyst + defensive-security-analyst + data-architect + analytics-data-engineer
↓
Execute concurrently → Merge results → Return
User: "Launch a new AI-powered product from scratch"
↓
Orchestrator: ai-lead-ops dynamically routes through many skills
↓
Execute with dynamic routing → Return
Each skill lives in <skill-name>/SKILL.md with YAML frontmatter (name, description triggers) and optional references/ depth material.
| Goal | What to do |
|---|---|
| Find a skill | Search this repo by keyword, or open .docs/SKILL_REGISTRY.md for the indexed catalog |
| Route work | Match user intent to a skill's description triggers; use ai-skill-manager for overlap, gaps, and portfolio hygiene |
| Run a workflow | Pick a template under .workflows/ (data pipeline, security ops, revenue ops, AI product dev, infra deployment) |
| Add a skill | Create <skill-name>/SKILL.md, then update registry and architecture docs (see Contributing) |
Agent clients (Cursor, Claude Code, etc.): prefer npx skills add daemon-blockint-tech/Agentic-Enteprises-Skill for installs into your project; alternatively clone this repo or copy <skill-name>/ folders into your agent skills path. Skills are discovered from SKILL.md files; no separate runtime is required beyond your agent host.
The OS still organizes work into seven layers. Skill count per layer has grown well beyond the original 86-skill baseline; treat layers as routing planes, not fixed headcount.
Layer 5: GOVERNANCE & ORCHESTRATION
↑
Layer 4: AI & ML OPERATIONS
↑
Layer 3: DATA & ANALYTICS
↑
Layer 2: INFRASTRUCTURE & SECURITY
↑
Layer 1: SOFTWARE ENGINEERING
↑
Layer 0: BUSINESS & OPERATIONS
↑
Layer -1: PHYSICAL INFRASTRUCTURE
Executable YAML in .workflows/ (v1.2.0, skills.sh install chains):
- Data Pipeline —
data-pipeline.yaml - Security Ops —
security-ops.yaml(CTI → hunt → offensive → D3FEND → compliance → IR) - Revenue Ops —
revenue-ops.yaml - AI Product Dev —
ai-product-dev.yaml(orchestrated) - Infra Deployment —
infra-deployment.yaml - Cloud & FinOps —
cloud-finops.yaml - Actuarial Credential —
actuarial-credential.yaml - AML Operations —
aml-operations.yaml
Compose manually: software dev (architecture → fullstack → frontend → UI → UX → support)
- build-validator: Quality gate at every layer
- ai-risk-governance: AI safety overlay
- compliance-engineer: Compliance overlay
- ai-memory-developer: Persistent context across sessions
| Document | Description |
|---|---|
| HIGH_LEVEL_DESIGN.md | High-level architecture and system design |
| ARCHITECTURE_GRAPH.md | Mermaid graph of skills and dependencies |
| ORCHESTRATION_ROUTING.md | Routing logic, execution modes, state management, error handling |
| SKILL_REGISTRY.md | Full 198-skill catalog (layer tables + skills.sh index) |
Note:
.docs/SKILL_REGISTRY.mdis generated from allSKILL.mdfiles (python3 scripts/generate_docs_registry.py). Architecture graph shows core relationships; full catalog is in the registry.
Each workflow includes package, install.chain_cursor (skills.sh CLI), and skills_sh_folder per step. Schema reference: .workflows/_workflow-schema.yaml.
| Workflow | Mode | Description | File |
|---|---|---|---|
| Data Pipeline | chain | Architecture → warehouse → analytics → science → BI → ops | .workflows/data-pipeline.yaml |
| Security Ops | chain | CTI → hunt → offensive → defensive → D3FEND → compliance → IR | .workflows/security-ops.yaml |
| Revenue Ops | chain | Counsel → deal ops → accounting → monetization → transactions | .workflows/revenue-ops.yaml |
| AI Product Dev | orchestrated | Research → safeguards → agentic AI → build → red team → gate | .workflows/ai-product-dev.yaml |
| Infra Deployment | chain | Infra → platform → K8s → CI/CD → DevSecOps → SRE → rollout | .workflows/infra-deployment.yaml |
| Cloud & FinOps | chain | Cloud architecture → security → compliance → FinOps → SRE | .workflows/cloud-finops.yaml |
| Actuarial Credential | chain | Foundations → exams → VEE → analyst → associate → actuary | .workflows/actuarial-credential.yaml |
| AML Operations | chain | AML program → monitoring → FIU → STR → false-positive tuning | .workflows/aml-operations.yaml |
| Layer | Representative skills |
|---|---|
| Governance & orchestration | build-validator, ai-risk-governance, ai-skill-manager, technical-program-manager, engineering-manager-* |
| AI & ML operations | ai-researcher, ai-engineer, ai-lead-ops, ai-redteam, prompt-engineer, ml-*-safeguards, ml-systems-engineer-rl-engineering |
| Data & analytics | data-architect, data-warehouse-engineer, analytics-data-engineer, data-scientist, bi-analyst, ontology-engineer |
| Infrastructure & security | infrastructure-engineer, platform-engineer, devops, devsecops, cybersecurity, offensive-security-analyst, defensive-security-analyst, D3FEND (d3fend-*) |
| Software engineering | senior-system-architecture, senior-fullstack-developer, ui-software-engineer, ux-software-engineer, support-engineer |
| Business & operations | business-consultant, commercial-counsel, deal-operations-administrator, product-management-monetization, customer-ops-specialist |
| Physical infrastructure | data-center-*, field-services-engineer, director-infrastructure-capex-accounting |
| Domain | Example skills |
|---|---|
| Actuarial & insurance | actuary, actuarial-analyst, actuarial-consulting, appointed-chief-actuary, associate-actuary, asset-liability-management, assumption-setting, advanced-*-actuarial-mathematics, property-casualty-insurance, life-health-insurance, pension-retirement-funds, ifrs, ab-testing-engineer |
| AML & financial crime | aml-cft, aml-compliance, financial-intelligence-unit, str-report, anti-false-positive-decision-making |
| Cloud, FinOps & platform | cloud-architect, cloud-engineer, cloud-security-engineer, cloud-economist, cloud-compliance-specialist, enterprise-cloud-architect, finops-analyst, vp-of-cloud, vp-of-infrastructure |
| SRE & resilience | site-reliability-engineer, sla-slo-engineer, bcm-disaster-recovery-specialist |
| Predictive & quantitative analytics | predictive-analytics, predictive-logistics-developer, quantitative-researcher, sentiment-analysis-engineer, sentiment-forecasting-engineer, operations-research-algorithm-developer |
| ML safeguards & agentic AI | ml-research-engineer-safeguards, ml-infrastructure-engineer-safeguards, privacy-research-engineer-safeguards, ai-adversarial-robustness-engineer, agentic-ai-developer, multi-agent-system-engineer |
| Security (expanded) | advanced-persistent-threat, cti-analyst, penetration-tester, red-team-specialist, web-pentester, network-pentester, cyber-resilience-engineer, enterprise-security-architect, vendor-cyber-risk-analyst, scada-ics-cyber-security-specialist, yara-rule-authoring, cryptographer-specialist, reverse-engineer, software-assurance-formal-methods-specialist, chief-information-security-officer, certified-information-systems-security-professional |
| Supply chain, IoT & edge | supply-chain-manager, iot-network-edge-engineer, robotics-automation-integration-engineer, wms-developer, geospatial-telematics-developer |
| Networking | cisco-certified-network-professional, sd-wan-engineer, network-backbone-architect, wireless-wifi-mobility-specialist |
For the full list, run find . -maxdepth 2 -name SKILL.md or browse .docs/SKILL_REGISTRY.md.
| Mode | Use When | Characteristics |
|---|---|---|
| Single | Quick, focused task | Low latency, single context |
| Chain | Multi-step workflow | Sequential, state passed between skills |
| Parallel | Independent tasks | Concurrent, results merged |
| Orchestrated | Complex, ambiguous tasks | Dynamic routing, AI manager |
Every skill output passes through build-validator:
- Architecture: System design is sound
- Security: No vulnerabilities
- Cost: Within budget
- Compliance: Meets regulations
- Performance: Meets targets
- Scalability: Can scale to load
User: "Build end-to-end analytics pipeline"
↓
Chain: data-architect → data-warehouse-engineer → analytics-data-engineer → data-scientist → bi-analyst
↓
Validation: build-validator after architect and scientist
↓
Return: Complete analytics pipeline with dashboard
User: "Full security audit with D3FEND response"
↓
Chain: offensive-security-analyst → defensive-security-analyst → D3FEND (7 skills) → compliance-engineer → incident-management-engineer
↓
Validation: build-validator after offensive and D3FEND
↓
Return: Complete security audit with defense plan
User: "Build AI-powered product from scratch"
↓
Orchestrator: ai-lead-ops
↓
Dynamic: business-model-researcher → ai-researcher → senior-system-architecture → data-architect → ai-engineer → senior-fullstack-developer → cybersecurity → compliance-engineer → build-validator
↓
Return: Complete AI product with security and compliance
- Create skill in
<skill-name>/SKILL.md(YAML frontmatter + triggers indescription) - Add references under
references/as needed - Regenerate registry:
python3 scripts/generate_docs_registry.py; update.docs/ARCHITECTURE_GRAPH.mdif routing relationships change - Run
build-validator(or portfolio checks viaai-skill-manager) to validate routing and overlap
Internal use only. Not for external distribution.
This is not a collection of prompts. This is an operating system for AI-native enterprises.