environment-setup-guide
Guide developers through setting up development environments with proper tools, dependencies, and configurations
obsidian-migration-deep-dive
Execute major Obsidian plugin rewrites and migration strategies. Use when migrating to or from Obsidian, performing major plugin rewrites, or re-platforming existing note systems to Obsidian. Trigger with phrases like "migrate to obsidian", "obsidian migration", "convert notes to obsidian", "obsi...
Full skill instructions
Comprehensive guide for migrating to Obsidian from other note-taking apps, or performing major plugin architecture rewrites.
| Type | Complexity | Duration | Risk |
|---|---|---|---|
| Single app import | Low | Hours | Low |
| Multi-source merge | Medium | Days | Medium |
| Plugin major rewrite | Medium | Weeks | Medium |
| Enterprise migration | High | Months | High |
// scripts/migration-assessment.ts
interface MigrationAssessment {
sourceSystem: string;
noteCount: number;
attachmentCount: number;
totalSize: number;
linkCount: number;
tagCount: number;
uniqueTags: string[];
folderStructure: string[];
issues: MigrationIssue[];
}
interface MigrationIssue {
type: 'encoding' | 'format' | 'link' | 'attachment' | 'metadata';
severity: 'warning' | 'error';
description: string;
affectedFiles: string[];
}
async function assessMigration(sourcePath: string): Promise<MigrationAssessment> {
const assessment: MigrationAssessment = {
sourceSystem: 'unknown',
noteCount: 0,
attachmentCount: 0,
totalSize: 0,
linkCount: 0,
tagCount: 0,
uniqueTags: [],
folderStructure: [],
issues: [],
};
// Scan source directory
// Count files, measure sizes
// Identify formats and potential issues
return assessment;
}
// Generate report
function generateAssessmentReport(assessment: MigrationAssessment): string {
return `
# Migration Assessment Report
## Source System: ${assessment.sourceSystem}
### Content Summary
- Notes: ${assessment.noteCount}
- Attachments: ${assessment.attachmentCount}
- Total Size: ${(assessment.totalSize / 1024 / 1024).toFixed(2)} MB
- Links: ${assessment.linkCount}
- Tags: ${assessment.tagCount} (${assessment.uniqueTags.length} unique)
### Folder Structure
${assessment.folderStructure.map(f => `- ${f}`).join('\n')}
### Issues Found
${assessment.issues.map(i => `- [${i.severity.toUpperCase()}] ${i.type}: ${i.description}`).join('\n')}
### Recommendations
${assessment.issues.length === 0 ? '- No issues found, proceed with migration' : '- Address issues before migration'}
`;
}
// scripts/converters/evernote.ts
import * as fs from 'fs';
import * as path from 'path';
import { parseStringPromise } from 'xml2js';
interface EvernoteNote {
title: string;
content: string;
created: string;
updated: string;
tags: string[];
attachments: EvernoteAttachment[];
}
interface EvernoteAttachment {
filename: string;
mime: string;
data: string; // base64
}
export async function convertEvernoteExport(
enexPath: string,
outputPath: string
): Promise<{ notes: number; attachments: number }> {
const content = fs.readFileSync(enexPath, 'utf-8');
const parsed = await parseStringPromise(content);
const notes = parsed['en-export']?.note || [];
let noteCount = 0;
let attachmentCount = 0;
for (const note of notes) {
const converted = convertEvernoteNote(note);
const fileName = sanitizeFileName(converted.title) + '.md';
const filePath = path.join(outputPath, fileName);
// Convert HTML content to Markdown
const markdown = convertHtmlToMarkdown(converted.content);
// Add frontmatter
const frontmatter = `---
title: ${converted.title}
created: ${converted.created}
updated: ${converted.updated}
tags: [${converted.tags.join(', ')}]
source: evernote
---
`;
fs.writeFileSync(filePath, frontmatter + markdown);
noteCount++;
// Handle attachments
for (const attachment of converted.attachments) {
const attachmentPath = path.join(outputPath, 'attachments', attachment.filename);
const data = Buffer.from(attachment.data, 'base64');
fs.writeFileSync(attachmentPath, data);
attachmentCount++;
}
}
return { notes: noteCount, attachments: attachmentCount };
}
function convertEvernoteNote(note: any): EvernoteNote {
return {
title: note.title?.[0] || 'Untitled',
content: note.content?.[0] || '',
created: formatDate(note.created?.[0]),
updated: formatDate(note.updated?.[0]),
tags: note.tag || [],
attachments: extractAttachments(note.resource || []),
};
}
// scripts/converters/notion.ts
export async function convertNotionExport(
notionPath: string,
outputPath: string
): Promise<{ notes: number; databases: number }> {
// Notion exports as nested folders with markdown/CSV
// Walk directory and convert
let noteCount = 0;
let databaseCount = 0;
// Implementation...
return { notes: noteCount, databases: databaseCount };
}
// scripts/converters/roam.ts
export async function convertRoamExport(
roamJsonPath: string,
outputPath: string
): Promise<{ pages: number; blocks: number }> {
const content = fs.readFileSync(roamJsonPath, 'utf-8');
const roamData = JSON.parse(content);
let pageCount = 0;
let blockCount = 0;
for (const page of roamData) {
const markdown = convertRoamPage(page);
const fileName = sanitizeFileName(page.title) + '.md';
fs.writeFileSync(path.join(outputPath, fileName), markdown);
pageCount++;
blockCount += countBlocks(page);
}
return { pages: pageCount, blocks: blockCount };
}
function convertRoamPage(page: any): string {
const lines: string[] = [`# ${page.title}`, ''];
if (page.children) {
for (const block of page.children) {
lines.push(...convertRoamBlock(block, 0));
}
}
return lines.join('\n');
}
function convertRoamBlock(block: any, depth: number): string[] {
const lines: string[] = [];
const indent = ' '.repeat(depth);
const content = convertRoamSyntax(block.string || '');
lines.push(`${indent}- ${content}`);
if (block.children) {
for (const child of block.children) {
lines.push(...convertRoamBlock(child, depth + 1));
}
}
return lines;
}
function convertRoamSyntax(text: string): string {
// Convert Roam-specific syntax to Obsidian
return text
.replace(/\[\[([^\]]+)\]\]/g, '[[$1]]') // Links same
.replace(/\(\(([^)]+)\)\)/g, '^$1') // Block refs to block IDs
.replace(/#\[\[([^\]]+)\]\]/g, '#$1') // Tag pages to tags
.replace(/{{embed: \[\[([^\]]+)\]\]}}/g, '![[$ 1]]'); // Embeds
}
// scripts/migrate-links.ts
import * as fs from 'fs';
import * as path from 'path';
import * as glob from 'glob';
interface LinkMapping {
original: string;
converted: string;
type: 'internal' | 'external' | 'attachment';
}
export class LinkMigrator {
private linkMappings: Map<string, LinkMapping> = new Map();
private orphanedLinks: string[] = [];
async buildLinkIndex(vaultPath: string): Promise<void> {
const files = glob.sync('**/*.md', { cwd: vaultPath });
for (const file of files) {
const baseName = path.basename(file, '.md');
this.linkMappings.set(baseName.toLowerCase(), {
original: baseName,
converted: baseName,
type: 'internal',
});
}
}
async migrateLinks(vaultPath: string): Promise<{
updated: number;
orphaned: string[];
}> {
const files = glob.sync('**/*.md', { cwd: vaultPath });
let updatedCount = 0;
for (const file of files) {
const filePath = path.join(vaultPath, file);
let content = fs.readFileSync(filePath, 'utf-8');
let modified = false;
// Find all wiki-style links
const linkRegex = /\[\[([^\]|]+)(\|[^\]]+)?\]\]/g;
let match;
while ((match = linkRegex.exec(content)) !== null) {
const originalLink = match[1];
const alias = match[2] || '';
const resolvedLink = this.resolveLink(originalLink);
if (resolvedLink !== originalLink) {
const newLink = `[[${resolvedLink}${alias}]]`;
content = content.replace(match[0], newLink);
modified = true;
}
}
if (modified) {
fs.writeFileSync(filePath, content);
updatedCount++;
}
}
return {
updated: updatedCount,
orphaned: this.orphanedLinks,
};
}
private resolveLink(link: string): string {
// Try exact match
const mapping = this.linkMappings.get(link.toLowerCase());
if (mapping) {
return mapping.converted;
}
// Try without path
const baseName = path.basename(link);
const baseMapping = this.linkMappings.get(baseName.toLowerCase());
if (baseMapping) {
return baseMapping.converted;
}
// Mark as orphaned
if (!this.orphanedLinks.includes(link)) {
this.orphanedLinks.push(link);
}
return link;
}
async createOrphanedLinksReport(vaultPath: string): Promise<void> {
const report = `# Orphaned Links Report
These links could not be resolved during migration:
${this.orphanedLinks.map(link => `- [[${link}]]`).join('\n')}
## Actions Needed
- Create missing notes
- Update or remove broken links
- Check for renamed files
`;
fs.writeFileSync(
path.join(vaultPath, '_migration', 'orphaned-links.md'),
report
);
}
}
// scripts/migrate.ts
import * as fs from 'fs';
import * as path from 'path';
import { convertEvernoteExport } from './converters/evernote';
import { convertNotionExport } from './converters/notion';
import { convertRoamExport } from './converters/roam';
import { LinkMigrator } from './migrate-links';
interface MigrationConfig {
source: {
type: 'evernote' | 'notion' | 'roam' | 'markdown';
path: string;
};
target: {
vaultPath: string;
createBackup: boolean;
};
options: {
preserveFolderStructure: boolean;
convertTags: boolean;
migrateAttachments: boolean;
fixLinks: boolean;
dryRun: boolean;
};
}
async function runMigration(config: MigrationConfig): Promise<void> {
console.log('Starting migration...');
console.log(`Source: ${config.source.type} from ${config.source.path}`);
console.log(`Target: ${config.target.vaultPath}`);
// Create backup if requested
if (config.target.createBackup && !config.options.dryRun) {
const backupPath = `${config.target.vaultPath}-backup-${Date.now()}`;
fs.cpSync(config.target.vaultPath, backupPath, { recursive: true });
console.log(`Backup created at: ${backupPath}`);
}
// Create migration folder for reports
const migrationFolder = path.join(config.target.vaultPath, '_migration');
if (!config.options.dryRun) {
fs.mkdirSync(migrationFolder, { recursive: true });
}
// Run appropriate converter
let result: { notes: number; [key: string]: number };
switch (config.source.type) {
case 'evernote':
result = await convertEvernoteExport(
config.source.path,
config.target.vaultPath
);
break;
case 'notion':
result = await convertNotionExport(
config.source.path,
config.target.vaultPath
);
break;
case 'roam':
result = await convertRoamExport(
config.source.path,
config.target.vaultPath
);
break;
default:
throw new Error(`Unsupported source type: ${config.source.type}`);
}
console.log(`Converted ${result.notes} notes`);
// Fix links if requested
if (config.options.fixLinks) {
const linkMigrator = new LinkMigrator();
await linkMigrator.buildLinkIndex(config.target.vaultPath);
const linkResult = await linkMigrator.migrateLinks(config.target.vaultPath);
console.log(`Updated links in ${linkResult.updated} files`);
console.log(`Found ${linkResult.orphaned.length} orphaned links`);
if (linkResult.orphaned.length > 0) {
await linkMigrator.createOrphanedLinksReport(config.target.vaultPath);
}
}
// Generate migration report
const report = generateMigrationReport(config, result);
if (!config.options.dryRun) {
fs.writeFileSync(
path.join(migrationFolder, 'migration-report.md'),
report
);
}
console.log('Migration complete!');
}
function generateMigrationReport(
config: MigrationConfig,
result: { notes: number; [key: string]: number }
): string {
return `# Migration Report
## Summary
- **Date:** ${new Date().toISOString()}
- **Source:** ${config.source.type}
- **Notes migrated:** ${result.notes}
## Configuration
\`\`\`json
${JSON.stringify(config, null, 2)}
\`\`\`
## Results
${Object.entries(result).map(([key, value]) => `- ${key}: ${value}`).join('\n')}
## Next Steps
1. Review migrated content
2. Check orphaned links report
3. Test in Obsidian
4. Remove _migration folder when satisfied
`;
}
// Run migration
const config: MigrationConfig = {
source: {
type: 'evernote',
path: '/path/to/export.enex',
},
target: {
vaultPath: '/path/to/obsidian/vault',
createBackup: true,
},
options: {
preserveFolderStructure: true,
convertTags: true,
migrateAttachments: true,
fixLinks: true,
dryRun: false,
},
};
runMigration(config).catch(console.error);
// For major plugin rewrites
interface PluginMigrationPlan {
currentVersion: string;
targetVersion: string;
phases: MigrationPhase[];
rollbackPlan: string;
}
interface MigrationPhase {
name: string;
description: string;
changes: string[];
breakingChanges: string[];
migrationSteps: string[];
}
const migrationPlan: PluginMigrationPlan = {
currentVersion: '1.x',
targetVersion: '2.0',
phases: [
{
name: 'Phase 1: Settings Migration',
description: 'Migrate settings to new format',
changes: [
'New settings schema',
'Split monolithic settings into categories',
],
breakingChanges: [
'Old settings format deprecated',
],
migrationSteps: [
'Load old settings on upgrade',
'Transform to new format',
'Save new settings',
'Backup old settings',
],
},
{
name: 'Phase 2: API Changes',
description: 'Update internal APIs',
changes: [
'New service-based architecture',
'Event-driven communication',
],
breakingChanges: [
'Direct vault access deprecated',
'Command IDs changed',
],
migrationSteps: [
'Update command registrations',
'Migrate to new services',
'Update event handlers',
],
},
],
rollbackPlan: 'Install previous version from GitHub releases',
};
| Issue | Cause | Solution |
|---|---|---|
| Encoding errors | Non-UTF8 content | Detect and convert encoding |
| Broken links | Renamed/deleted files | Generate orphaned links report |
| Missing attachments | Export incomplete | Re-export with attachments |
| Duplicate files | Same name different folders | Add path prefix |
# Install dependencies
npm install xml2js glob
# Run migration
npx ts-node scripts/migrate.ts
# Dry run first
MIGRATION_DRY_RUN=true npx ts-node scripts/migrate.ts
## Post-Migration Checklist
- [ ] Open vault in Obsidian
- [ ] Check random sample of notes (10-20)
- [ ] Verify links resolve correctly
- [ ] Check attachments display
- [ ] Verify tags imported
- [ ] Test search functionality
- [ ] Check folder structure
- [ ] Review orphaned links report
- [ ] Delete _migration folder
- [ ] Update any external integrations
Migration complete! You now have comprehensive Obsidian plugin development skills.
Guide developers through setting up development environments with proper tools, dependencies, and configurations
Implement Exa lint rules, policy enforcement, and automated guardrails. Use when setting up code quality rules for Exa integrations, implementing pre-commit hooks, or configuring CI policy checks for Exa best practices. Trigger with phrases like "exa policy", "exa lint", "exa guardrails", "exa be...
Configure CI/CD pipelines for Documenso integrations. Use when setting up automated testing, deployment pipelines, or continuous integration for Documenso projects. Trigger with phrases like "documenso CI", "documenso GitHub Actions", "documenso pipeline", "documenso automated testing".
Build check model availability and implement fallback chains. Use when building resilient systems or handling model outages. Trigger with phrases like 'openrouter availability', 'openrouter fallback', 'openrouter model down', 'openrouter health check'.
Configure Langfuse enterprise organization management and access control. Use when implementing team access controls, configuring organization settings, or setting up role-based permissions for Langfuse projects. Trigger with phrases like "langfuse RBAC", "langfuse teams", "langfuse organization"...
Execute compliance and security auditing for Cursor usage. Triggers on "cursor compliance", "cursor audit", "cursor security review", "cursor soc2", "cursor gdpr". Use when analyzing or auditing cursor compliance audit. Trigger with phrases like "cursor compliance audit", "cursor audit", "cursor".
Configure Apollo.io CI/CD integration. Use when setting up automated testing, continuous integration, or deployment pipelines for Apollo integrations. Trigger with phrases like "apollo ci", "apollo github actions", "apollo pipeline", "apollo ci/cd", "apollo automated tests".
Create robust Python automation with full logging and safety checks. Use when tasks need complex data processing, authenticated API work, conditional file operations, or error handling beyond simple shell commands.
Configure PostHog across development, staging, and production environments. Use when setting up multi-environment deployments, configuring per-environment secrets, or implementing environment-specific PostHog configurations. Trigger with phrases like "posthog environments", "posthog staging", "po...
Expert in designing and building autonomous AI agents. Masters tool use, memory systems, planning strategies, and multi-agent orchestration. Use when: build agent, AI agent, autonomous agent, tool use, function calling.
Validate insecure deserialization checker operations. Auto-activating skill for Security Fundamentals. Triggers on: insecure deserialization checker, insecure deserialization checker Part of the Security Fundamentals skill category. Use when working with insecure deserialization checker functiona...
Manage path traversal finder operations. Auto-activating skill for Security Fundamentals. Triggers on: path traversal finder, path traversal finder Part of the Security Fundamentals skill category. Use when working with path traversal finder functionality. Trigger with phrases like "path traversa...
Generate, edit, and beat-sync AI video with leading models in one workspace.
The world's fastest calendar for remote work
Transform Your Design with AI Designer by ImgCreator.ai
Revolutionizing Video Production with AI-Powered Creativity
Extend an image past the frame and let AI fill the new aspect ratio.
Discover your celebrity doppelgänger with StarByFace!
ChainClarity explains 700+ crypto whitepapers in plain English, with layered summaries, comparisons, research tools, alerts, and a $4.99 Pro plan.
Opus.ai: Revolutionize Your Web Experience