메뉴
Neural network training and deployment in Flow Nexus cloud. Use for distributed ML training, model inference, and neural network lifecycle management.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/majiayu000/claude-skill-registry --skill cloud-neural명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
LLM token logprobs and calibration. Per-decision confidence, ECE, Brier, reliability diagrams, low-confidence triage.
Analyze LLM token logprobs and calibration. Use for per-decision confidence, ECE, Brier scores, reliability diagrams, and low-confidence triage.
回顾最近 N 天的 Claude Code 使用记录——扫描原始会话数据，按主题分组汇总"我都做了什么"，并从个人操作系统视角输出模式、风险与增删建议。当用户说 /recap、"看看我这几天做了什么"、"回顾一下我最近的会话"、"这两天我用 claude 干了啥"、"活动回顾" 时使用。
SOC 직업 분류 기준
SKILL.md 표시 중
| name | cloud-neural |
| description | Neural network training and deployment in Flow Nexus cloud. Use for distributed ML training, model inference, and neural network lifecycle management. |
| version | 1.0.0 |
| category | cloud |
| type | skill |
| capabilities | ["neural_network_training","distributed_training","model_inference","cluster_management","template_deployment","model_validation"] |
| tools | ["mcp__flow-nexus__neural_train","mcp__flow-nexus__neural_predict","mcp__flow-nexus__neural_cluster_init","mcp__flow-nexus__neural_node_deploy","mcp__flow-nexus__neural_cluster_connect","mcp__flow-nexus__neural_train_distributed","mcp__flow-nexus__neural_cluster_status","mcp__flow-nexus__neural_predict_distributed","mcp__flow-nexus__neural_cluster_terminate","mcp__flow-nexus__neural_list_templates","mcp__flow-nexus__neural_deploy_template","mcp__flow-nexus__neural_training_status","mcp__flow-nexus__neural_list_models","mcp__flow-nexus__neural_validation_workflow","mcp__flow-nexus__neural_publish_template","mcp__flow-nexus__neural_rate_template","mcp__flow-nexus__neural_performance_benchmark"] |
| related_skills | ["cloud-swarm","cloud-sandbox","cloud-workflow"] |
Train, deploy, and manage neural networks at scale using Flow Nexus cloud-powered distributed computing.
// Train a basic neural network
mcp__flow-nexus__neural_train({
config: {
architecture: {
type: "feedforward",
layers: [
{ type: "dense", units: 128, activation: "relu" },
{ type: "dropout", rate: 0.2 },
{ type: "dense", units: 10, activation: "softmax" }
]
},
training: { epochs: 100, batch_size: 32, learning_rate: 0.001 }
},
tier: "small"
})
// Run inference
mcp__flow-nexus__neural_predict({
model_id: "trained_model_id",
input: [[0.5, 0.3, 0.2]]
})
flow-nexus configured| Type | Use Case |
|---|---|
| Feedforward | Classification, regression |
| LSTM/RNN | Time series, NLP sequences |
| Transformer | Advanced NLP, multimodal |
| CNN | Computer vision, image processing |
| GAN | Data generation, augmentation |
| Autoencoder | Dimensionality reduction, anomaly detection |
| Tier | Resources | Cost |
|---|---|---|
nano | Minimal, quick tests | Low |
mini | Small models | Low |
small | Standard training | Medium |
medium | Large models | High |
large | Production scale | Highest |
mcp__flow-nexus__neural_train({
config: {
architecture: {
type: "feedforward", // lstm, gan, autoencoder, transformer
layers: [
{ type: "dense", units: 128, activation: "relu" },
{ type: "dropout", rate: 0.2 },
{ type: "dense", units: 10, activation: "softmax" }
]
},
training: {
epochs: 100,
batch_size: 32,
learning_rate: 0.001,
optimizer: "adam"
},
divergent: {
enabled: false,
pattern: "lateral", // quantum, chaotic, associative, evolutionary
factor: 0.1
}
},
tier: "small", // nano, mini, small, medium, large
user_id: "user_id"
})
// Initialize distributed cluster
mcp__flow-nexus__neural_cluster_init({
name: "training-cluster",
architecture: "transformer", // transformer, cnn, rnn, gnn, hybrid
topology: "mesh", // mesh, ring, star, hierarchical
consensus: "proof-of-learning",
daaEnabled: true,
wasmOptimization: true
})
// Deploy worker nodes
mcp__flow-nexus__neural_node_deploy({
cluster_id: "cluster_id",
node_type: "worker", // worker, parameter_server, aggregator, validator
model: "base", // base, large, xl, custom
capabilities: ["training", "inference"],
autonomy: 0.8
})
// Connect nodes based on topology
mcp__flow-nexus__neural_cluster_connect({
cluster_id: "cluster_id",
topology: "mesh"
})
// Start distributed training
mcp__flow-nexus__neural_train_distributed({
cluster_id: "cluster_id",
dataset: "dataset_id",
epochs: 10,
batch_size: 32,
learning_rate: 0.001,
optimizer: "adam",
federated: false
})
// Check cluster status
mcp__flow-nexus__neural_cluster_status({ cluster_id: "cluster_id" })
// Terminate when done
mcp__flow-nexus__neural_cluster_terminate({ cluster_id: "cluster_id" })
// Single-node inference
mcp__flow-nexus__neural_predict({
model_id: "model_id",
input: [[0.5, 0.3, 0.2]],
user_id: "user_id"
})
// Distributed inference
mcp__flow-nexus__neural_predict_distributed({
cluster_id: "cluster_id",
input_data: "[0.5, 0.3, 0.2]",
aggregation: "mean" // mean, majority, weighted, ensemble
})
// List templates
mcp__flow-nexus__neural_list_templates({
category: "classification", // timeseries, regression, nlp, vision, anomaly, generative, reinforcement, custom
tier: "free", // free, paid
search: "sentiment",
limit: 20
})
// Deploy template
mcp__flow-nexus__neural_deploy_template({
template_id: "template_id",
custom_config: { epochs: 50 },
user_id: "user_id"
})
// Publish your model as template
mcp__flow-nexus__neural_publish_template({
model_id: "model_id",
name: "Sentiment Analyzer",
description: "LSTM-based sentiment analysis model",
category: "nlp",
price: 0,
user_id: "user_id"
})
// Rate a template
mcp__flow-nexus__neural_rate_template({
template_id: "template_id",
rating: 5,
review: "Excellent model, fast and accurate",
user_id: "user_id"
})
// List user models
mcp__flow-nexus__neural_list_models({
user_id: "user_id",
include_public: false
})
// Check training status
mcp__flow-nexus__neural_training_status({ job_id: "job_id" })
// Create validation workflow
mcp__flow-nexus__neural_validation_workflow({
model_id: "model_id",
validation_type: "comprehensive", // performance, accuracy, robustness, comprehensive
user_id: "user_id"
})
// Run performance benchmarks
mcp__flow-nexus__neural_performance_benchmark({
model_id: "model_id",
benchmark_type: "comprehensive" // inference, throughput, memory, comprehensive
})
// Train a feedforward classifier
const trainingJob = await mcp__flow-nexus__neural_train({
config: {
architecture: {
type: "feedforward",
layers: [
{ type: "dense", units: 256, activation: "relu" },
{ type: "batch_norm" },
{ type: "dropout", rate: 0.3 },
{ type: "dense", units: 128, activation: "relu" },
{ type: "dropout", rate: 0.2 },
{ type: "dense", units: 10, activation: "softmax" }
]
},
training: {
epochs: 100,
batch_size: 64,
learning_rate: 0.001,
optimizer: "adam"
}
},
tier: "small"
});
// Monitor training
const status = await mcp__flow-nexus__neural_training_status({
job_id: trainingJob.job_id
});
console.log(`Epoch: ${status.current_epoch}, Loss: ${status.loss}`);
// Run inference on trained model
const prediction = await mcp__flow-nexus__neural_predict({
model_id: trainingJob.model_id,
input: [[0.1, 0.2, 0.3, 0.4, 0.5]]
});
// Initialize distributed cluster
const cluster = await mcp__flow-nexus__neural_cluster_init({
name: "transformer-cluster",
architecture: "transformer",
topology: "mesh",
consensus: "proof-of-learning",
daaEnabled: true,
wasmOptimization: true
});
// Deploy 4 worker nodes
for (let i = 0; i < 4; i++) {
await mcp__flow-nexus__neural_node_deploy({
cluster_id: cluster.cluster_id,
node_type: "worker",
model: "large",
capabilities: ["training", "inference"]
});
}
// Deploy parameter server
await mcp__flow-nexus__neural_node_deploy({
cluster_id: cluster.cluster_id,
node_type: "parameter_server",
model: "base"
});
// Connect nodes
await mcp__flow-nexus__neural_cluster_connect({
cluster_id: cluster.cluster_id
});
// Start distributed training
await mcp__flow-nexus__neural_train_distributed({
cluster_id: cluster.cluster_id,
dataset: "large_nlp_dataset",
epochs: 50,
batch_size: 128,
learning_rate: 0.0001,
optimizer: "adam"
});
// Monitor and validate
const clusterStatus = await mcp__flow-nexus__neural_cluster_status({
cluster_id: cluster.cluster_id
});
// Cleanup
await mcp__flow-nexus__neural_cluster_terminate({
cluster_id: cluster.cluster_id
});
// Find NLP templates
const templates = await mcp__flow-nexus__neural_list_templates({
category: "nlp",
tier: "free",
search: "sentiment"
});
// Deploy the best-rated template
const deployment = await mcp__flow-nexus__neural_deploy_template({
template_id: templates.templates[0].id,
custom_config: {
epochs: 25,
learning_rate: 0.0005
}
});
// Validate model performance
await mcp__flow-nexus__neural_validation_workflow({
model_id: deployment.model_id,
validation_type: "comprehensive"
});
// Benchmark performance
const benchmark = await mcp__flow-nexus__neural_performance_benchmark({
model_id: deployment.model_id,
benchmark_type: "comprehensive"
});
console.log(`Inference latency: ${benchmark.inference_latency_ms}ms`);
nano or mini tier for testing, scale up for production| Error | Cause | Solution |
|---|---|---|
training_failed | Invalid architecture config | Verify layer compatibility and types |
cluster_init_failed | Invalid topology or architecture | Check supported combinations |
insufficient_credits | Training tier exceeds balance | Reduce tier or add credits |
model_not_found | Invalid model_id | Use neural_list_models to verify |
node_deploy_failed | Cluster capacity reached | Terminate unused nodes |
// Deploy neural agent in swarm
await mcp__flow-nexus__agent_spawn({
type: "analyst",
name: "ML Analyst",
capabilities: ["neural_training", "model_evaluation"]
});
// ML pipeline workflow
await mcp__flow-nexus__workflow_create({
name: "ML Training Pipeline",
steps: [
{ id: "preprocess", action: "data_prep" },
{ id: "train", action: "neural_train", depends: ["preprocess"] },
{ id: "validate", action: "neural_validate", depends: ["train"] },
{ id: "deploy", action: "neural_deploy", depends: ["validate"] }
]
});