AGENT SKILL

vastai-migration-deep-dive

Execute Vast.ai major re-architecture and migration strategies with strangler fig pattern. Use when migrating to or from Vast.ai, performing major version upgrades, or re-platforming existing integrations to Vast.ai. Trigger with phrases like "migrate vastai", "vastai migration", "switch to vastai", "vastai replatform", "vastai upgrade major".

License 许可
MIT
最近更新
2026-11-15
Stars
★ 2.0k
Forks
233

01 前言 Front Matter

name
vastai-migration-deep-dive
author
Jeremy Longshore <jeremy@intentsolutions.io>
license
MIT
version
1.0.0
description
Execute Vast.ai major re-architecture and migration strategies with strangler fig pattern. Use when migrating to or from Vast.ai, performing major version upgrades, or re-platforming existing integrations to Vast.ai. Trigger with phrases like "migrate vastai", "vastai migration", "switch to vastai", "vastai replatform", "vastai upgrade major".
allowed-tools
Read, Write, Edit, Bash(npm:*), Bash(node:*), Bash(kubectl:*)

03 SKILL.md 正文

Vast.ai Migration Deep Dive

Overview

Comprehensive guide for migrating to or from Vast.ai, or major version upgrades.

Prerequisites

  • Current system documentation
  • Vast.ai SDK installed
  • Feature flag infrastructure
  • Rollback strategy tested

Migration Types

Type Complexity Duration Risk
Fresh install Low Days Low
From competitor Medium Weeks Medium
Major version Medium Weeks Medium
Full replatform High Months High

Pre-Migration Assessment

Step 1: Current State Analysis

# Document current implementation
find . -name "*.ts" -o -name "*.py" | xargs grep -l "vastai" > vastai-files.txt

# Count integration points
wc -l vastai-files.txt

# Identify dependencies
npm list | grep vastai
pip freeze | grep vastai

Step 2: Data Inventory

interface MigrationInventory {
  dataTypes: string[];
  recordCounts: Record<string, number>;
  dependencies: string[];
  integrationPoints: string[];
  customizations: string[];
}

async function assessVast.aiMigration(): Promise<MigrationInventory> {
  return {
    dataTypes: await getDataTypes(),
    recordCounts: await getRecordCounts(),
    dependencies: await analyzeDependencies(),
    integrationPoints: await findIntegrationPoints(),
    customizations: await documentCustomizations(),
  };
}

Migration Strategy: Strangler Fig Pattern

Phase 1: Parallel Run
┌─────────────┐     ┌─────────────┐
│   Old       │     │   New       │
│   System    │ ──▶ │  Vast.ai   │
│   (100%)    │     │   (0%)      │
└─────────────┘     └─────────────┘

Phase 2: Gradual Shift
┌─────────────┐     ┌─────────────┐
│   Old       │     │   New       │
│   (50%)     │ ──▶ │   (50%)     │
└─────────────┘     └─────────────┘

Phase 3: Complete
┌─────────────┐     ┌─────────────┐
│   Old       │     │   New       │
│   (0%)      │ ──▶ │   (100%)    │
└─────────────┘     └─────────────┘

Implementation Plan

Phase 1: Setup (Week 1-2)

# Install Vast.ai SDK
npm install @vastai/sdk

# Configure credentials
cp .env.example .env.vastai
# Edit with new credentials

# Verify connectivity
node -e "require('@vastai/sdk').ping()"

Phase 2: Adapter Layer (Week 3-4)

// src/adapters/vastai.ts
interface ServiceAdapter {
  create(data: CreateInput): Promise<Resource>;
  read(id: string): Promise<Resource>;
  update(id: string, data: UpdateInput): Promise<Resource>;
  delete(id: string): Promise<void>;
}

class Vast.aiAdapter implements ServiceAdapter {
  async create(data: CreateInput): Promise<Resource> {
    const vastaiData = this.transform(data);
    return vastaiClient.create(vastaiData);
  }

  private transform(data: CreateInput): Vast.aiInput {
    // Map from old format to Vast.ai format
  }
}

Phase 3: Data Migration (Week 5-6)

async function migrateVast.aiData(): Promise<MigrationResult> {
  const batchSize = 100;
  let processed = 0;
  let errors: MigrationError[] = [];

  for await (const batch of oldSystem.iterateBatches(batchSize)) {
    try {
      const transformed = batch.map(transform);
      await vastaiClient.batchCreate(transformed);
      processed += batch.length;
    } catch (error) {
      errors.push({ batch, error });
    }

    // Progress update
    console.log(`Migrated ${processed} records`);
  }

  return { processed, errors };
}

Phase 4: Traffic Shift (Week 7-8)

// Feature flag controlled traffic split
function getServiceAdapter(): ServiceAdapter {
  const vastaiPercentage = getFeatureFlag('vastai_migration_percentage');

  if (Math.random() * 100 < vastaiPercentage) {
    return new Vast.aiAdapter();
  }

  return new LegacyAdapter();
}

Rollback Plan

# Immediate rollback
kubectl set env deployment/app VASTAI_ENABLED=false
kubectl rollout restart deployment/app

# Data rollback (if needed)
./scripts/restore-from-backup.sh --date YYYY-MM-DD

# Verify rollback
curl https://app.yourcompany.com/health | jq '.services.vastai'

Post-Migration Validation

async function validateVast.aiMigration(): Promise<ValidationReport> {
  const checks = [
    { name: 'Data count match', fn: checkDataCounts },
    { name: 'API functionality', fn: checkApiFunctionality },
    { name: 'Performance baseline', fn: checkPerformance },
    { name: 'Error rates', fn: checkErrorRates },
  ];

  const results = await Promise.all(
    checks.map(async c => ({ name: c.name, result: await c.fn() }))
  );

  return { checks: results, passed: results.every(r => r.result.success) };
}

Instructions

Step 1: Assess Current State

Document existing implementation and data inventory.

Step 2: Build Adapter Layer

Create abstraction layer for gradual migration.

Step 3: Migrate Data

Run batch data migration with error handling.

Step 4: Shift Traffic

Gradually route traffic to new Vast.ai integration.

Output

  • Migration assessment complete
  • Adapter layer implemented
  • Data migrated successfully
  • Traffic fully shifted to Vast.ai

Error Handling

Issue Cause Solution
Data mismatch Transform errors Validate transform logic
Performance drop No caching Add caching layer
Rollback triggered Errors spiked Reduce traffic percentage
Validation failed Missing data Check batch processing

Examples

Quick Migration Status

const status = await validateVast.aiMigration();
console.log(`Migration ${status.passed ? 'PASSED' : 'FAILED'}`);
status.checks.forEach(c => console.log(`  ${c.name}: ${c.result.success}`));

Resources

Flagship+ Skills

For advanced troubleshooting, see vastai-advanced-troubleshooting.

04 同仓库更多 Skills

Dicklesworthstone
Dicklesworthstone/pi_agent_rust

theme-factory

Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.

视觉设计
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

web-artifacts-builder

Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.

前端开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

webapp-testing

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

测试与质量前端开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

neuropixels-analysis

Neuropixels neural recording analysis. Load SpikeGLX/OpenEphys data, preprocess, motion correction, Kilosort4 spike sorting, quality metrics, Allen/IBL curation, AI-assisted visual analysis, for Neuropixels 1.0/2.0 extracellular electrophysiology. Use when working with neural recordings, spike sorting, extracellular electrophysiology, or when the user mentions Neuropixels, SpikeGLX, Open Ephys, Kilosort, quality metrics, or unit curation.

★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Skill与智能体开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

notebooklm

Use this skill to query your Google NotebookLM notebooks directly from Claude Code for source-grounded, citation-backed answers from Gemini. Browser automation, library management, persistent auth. Drastically reduced hallucinations through document-only responses.

Skill与智能体开发
★ 2.0k