convez.ai
SkillProdutividadeIntermediário
ArdurAI

God-Level Skill Suite

ArdurAI/god-skill-suite

52 skills battle-tested para IA código sem alucinações

Ver no GitHub
InicialPluginScore · de 0 a 100

Conjunto de 52 skills de produção para Claude Code, Cursor, Windsurf e outras IDEs com IA. Cada skill reposiciona o modelo como um especialista em engenharia de software, eliminando alucinações e entregando código verificado. Cobre arquitetura, DevOps, segurança, frontend, backend e cross-domain mastery.

Instalação

$npx convez-ai ArdurAI/god-skill-suite

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
3
Forks
0
Acompanhando
0
Saúde
62%
Último commit
há 5 meses
Versão
v1.0.0
Issues abertas
1
Contribuidores
1

Casos de uso

  • Aumentar qualidade de código gerado por Claude Code, Cursor e Windsurf
  • Eliminar alucinações em APIs, sintaxe e decisões arquiteturais
  • Padronizar expertise de engenharia em equipes de desenvolvimento

Veredito

Suite madura e prática que resolve um problema real: alucinações em assistentes de código. As 52 skills são rigorosamente testadas e seu design anti-hallucination (forçar explicitação de incerteza) é uma abordagem sólida. Recomendado para times que usam Claude Code ou Cursor em produção.

README

A tradução para o português deste README ainda está na fila.

🌌 God-Level Skill Suite

52 Battle-Tested, Production-Grade AI Skills. Covering every dimension of modern software engineering. Zero Hallucination.

License: MIT PyPI version Skills Tested OS: Global

Built natively for Claude Code, Cursor, Windsurf, Continue, Gemini CLI, and more.

⭐ If this makes your AI coding sessions dramatically better, please star the repo! It helps other engineers discover the suite and motivates us to keep adding world-class skills.

📢 Share it with your team, in your Discord, or on Twitter/X — the more engineers use these skills, the more battle-tested they become.

🤝 Want to contribute a skill? Read the Contribution Guide — even a single high-quality skill makes a huge impact.


🎯 The Philosophy Grid

🛡️ Zero Hallucination⚔️ Researcher-Warrior🌐 Cross-Domain Mastery
No guessing. Strict anti-hallucination protocols force the model to invoke explicit uncertainty clauses rather than invent syntax or fabricate API behaviour.No passivity. Every skill adopts the persona of a relentless engineer who traces bugs to their fundamental root and deeply interrogates every architectural decision.No role siloing. A frontend engineer must understand Kubernetes networking. A DevOps engineer must understand OAuth flows. Lines fade deliberately.

⚡ How It Works


🚀 One-Minute Deployment

curl -fsSL https://raw.githubusercontent.com/gnanirahulnutakki/god-skill-suite/main/installer/install.sh | bash
irm https://raw.githubusercontent.com/gnanirahulnutakki/god-skill-suite/main/installer/install.ps1 | iex

Install safely from PyPI for global CLI access across your system.

# With uv (fastest)
uv tool install god-skill-suite && god-skills install

# With pipx
pipx install god-skill-suite && god-skills install

# Direct, no install required (run from cloned repo)
python installer/install.py

📚 The 52-Skill Registry

[!IMPORTANT] Always load god-meta-conductor first. It is the baseline operating system — establishing Zero-Hallucination rules and scope-locking before any domain skill activates.

SkillDescription
god-meta-conductorMaster orchestrator — anti-hallucination, cross-domain thinking, researcher-warrior identity
god-dev-coreEnd-to-end software engineering: DSA, OOP, SOLID/DRY/YAGNI, design patterns, clean code
god-dev-researchResearch methodology: finding papers, novelty checking, arXiv/ACM/IEEE, GitHub research
god-dev-builderBuilding production systems: architecture decisions, scaffolding, iterative delivery
god-dev-codebaseWorking with existing codebases: reading, debugging, refactoring, PR reviews
god-research-reviewDeep research review: literature synthesis, critical analysis, experimental design
SkillDescription
god-security-coreSTRIDE, OWASP Top 10, zero-trust, cryptography, SLSA, SBOM, Vault, container hardening
god-security-cloudCloud attack kill chain, privilege escalation, GuardDuty/Defender/SCC, CIEM
god-auth-protocolsOAuth 2.0 PKCE, OIDC, SAML 2.0, JWT, mTLS, SPIFFE/SPIRE, WebAuthn/passkeys
god-iam-awsAWS IAM: policies, roles, SCPs, permission boundaries, IRSA, IAM Access Analyzer
god-iam-azureAzure Entra ID: RBAC, Managed Identities, PIM, Conditional Access
god-iam-gcpGCP IAM: resource hierarchy, Workload Identity, Organization Policy, IAM Deny
SkillDescription
god-devops-coreCI/CD pipelines, GitOps, Docker, Helm, deployment strategies
god-devops-kubernetesKubernetes deep dive: pods/services/ingress/RBAC/HPA/PDB/NetworkPolicy
god-linux-masteryLinux internals, kernel/process/memory, eBPF/bpftrace, cgroups v2, systemd
god-containers-advancedOCI spec, OverlayFS, BuildKit, multi-arch, rootless containers, gVisor/Kata
god-infra-as-codeTerraform (advanced HCL, state, modules), Pulumi, AWS CDK, Crossplane
SkillDescription
god-web3-blockchainSolidity, EVM internals, Rust (Solana), smart contract security, DeFi, gas optimization
god-database-masteryPostgreSQL internals, EXPLAIN ANALYZE, indexing, sharding, replication, MongoDB, Redis
god-data-engineeringSpark, Flink, Airflow/Prefect, dbt, data modeling, lakehouse, streaming pipelines
god-data-cleaningData preprocessing, feature engineering, outlier detection, imputation, pipeline design
god-message-streamingKafka internals (ISR, EOS, Streams), RabbitMQ, SQS/SNS, NATS JetStream, Pulsar
SkillDescription
god-frontend-masteryReact 18+, Next.js App Router, TypeScript, state management, Core Web Vitals, WCAG 2.2
god-ui-ux-designInterface design, HSL color theory, 8pt grids, accessibility WCAG 2.2, Spring physics
god-mobile-awarenessiOS/SwiftUI, Android/Compose, React Native, Flutter, PWA, push notifications
SkillDescription
god-ai-promptingSystem prompts, chain-of-thought, few-shot, constitutional AI, evaluation
god-ai-architectRAG, agents, tool use, multimodal, inference optimization, responsible AI
god-llm-sdkClaude SDK, OpenAI SDK, LangChain, LlamaIndex, streaming, function calling
god-mlops-coreML pipelines, model versioning (MLflow/W&B), serving, drift detection, A/B testing
god-mlops-llmLLM fine-tuning (LoRA/QLoRA), RLHF, evaluation (RAGAS/DeepEval), LLMOps
god-ml-data-trainingData labeling, training loops, loss functions, distributed training, checkpointing
SkillDescription
god-serverless-architectureAWS Lambda, DynamoDB single-table, API Gateway, Step Functions, idempotency
god-systems-designCAP/PACELC, consensus (Raft/Paxos), consistent hashing, CQRS, Event Sourcing, Saga
god-architecture-patternsMicroservices, DDD, hexagonal/clean/onion, strangler fig, event-driven, ADRs
god-api-designREST (Richardson Maturity), GraphQL (N+1, DataLoader), gRPC, API versioning
SkillDescription
god-observabilityOpenTelemetry, Prometheus, Grafana, distributed tracing (Jaeger/Zipkin), alerting
god-sre-reliabilitySLOs/SLIs/error budgets, toil reduction, capacity planning, incident response, GameDay
god-performance-engineeringCPU/memory profiling, flame graphs, JVM tuning, Go pprof, k6/Gatling
god-platform-engineeringIDP design, Backstage, golden paths, developer portals, paved roads
SkillDescription
god-search-engineeringElasticsearch/OpenSearch internals, BM25 tuning, analyzer pipelines, relevance scoring
god-vector-databasesPinecone, Qdrant, Weaviate, pgvector, Chroma, FAISS — ANN algorithms, HNSW, hybrid search
god-edge-computingCloudflare Workers (KV/R2/Durable Objects), Lambda@Edge, Fastly, Vercel Edge
god-networkingTCP internals (BBR/CUBIC), DNS, TLS 1.3, VPC design, BGP, eBPF, troubleshooting
SkillDescription
god-compliance-governanceSOC 2 Type II, HIPAA, PCI-DSS v4.0, GDPR, ISO 27001:2022, NIST 800-53
god-cost-engineeringFinOps, cloud cost optimization, reserved instances, rightsizing, cost attribution
god-project-managementAgile/Scrum, Kanban, OKRs, RICE scoring, DORA metrics, blameless postmortems
god-testing-masteryUnit/integration/E2E/contract/load/chaos/fuzz/mutation testing — full test pyramid
god-git-workflowGit internals, rebasing, bisect, reflog, monorepo, hooks, conventional commits
god-devex-toolingVS Code/JetBrains/Neovim mastery, tmux, fzf, ripgrep, debuggers, linters
god-tech-supportFull-stack troubleshooting: reading stack traces, log analysis, network debugging
god-tech-supportKubernetes cluster triage, network debugging, systematic root-cause analysis

🧪 Validated Ecosystem

Every skill ships with a Prompt Library (prompts/examples.md) and an adversarial test suite (tests.json) that runs against local Ollama models or frontier APIs to empirically prove Zero-Hallucination compliance.

# Run the full 52-skill test suite locally (requires Ollama)
python scripts/evaluate_skills.py --provider ollama --model qwen2.5:7b --output docs/TEST_RESULTS.md

→ See docs/TESTING.md for the full framework guide and docs/TEST_RESULTS.md for latest results.


📄 License

MIT License — Designed for builders who refuse to guess.

Ver o original no GitHub

Bom para

Relacionados