profclaw 2.0.0 → 2.3.0

package/dist/cli/commands/import.js

@@ -0,0 +1,526 @@
+/**
+ * Import/Migration CLI Command
+ *
+ * Imports data from other AI assistants into profClaw:
+ *   - OpenClaw: memories, skills, conversation history
+ *   - ChatGPT: exported conversations (JSON)
+ *   - Aider: chat history
+ *
+ * Usage:
+ *   profclaw import --from openclaw [--path ~/.openclaw]
+ *   profclaw import --from chatgpt --file conversations.json
+ *   profclaw import --from aider [--path .]
+ */
+import { readFile, readdir, stat } from 'node:fs/promises';
+import { join, basename } from 'node:path';
+import { homedir } from 'node:os';
+import { Command } from 'commander';
+import { createContextualLogger } from '../../utils/logger.js';
+const log = createContextualLogger('Import');
+// OpenClaw Import
+/**
+ * Discover OpenClaw data directory.
+ * Checks: ~/.openclaw, OPENCLAW_HOME env, provided path.
+ */
+function findOpenClawPath(providedPath) {
+    if (providedPath)
+        return providedPath;
+    if (process.env['OPENCLAW_HOME'])
+        return process.env['OPENCLAW_HOME'];
+    return join(homedir(), '.openclaw');
+}
+/**
+ * Parse an OpenClaw MEMORY.md index file to find memory file paths.
+ */
+function parseMemoryIndex(content, basePath) {
+    const paths = [];
+    const linkPattern = /\[([^\]]+)\]\(([^)]+)\)/g;
+    let match;
+    while ((match = linkPattern.exec(content)) !== null) {
+        const relPath = match[2];
+        if (relPath.endsWith('.md') && !relPath.startsWith('http')) {
+            paths.push(join(basePath, relPath));
+        }
+    }
+    return paths;
+}
+/**
+ * Parse an OpenClaw memory file (YAML frontmatter + markdown body).
+ */
+function parseOpenClawMemory(content, filePath) {
+    // Parse YAML frontmatter
+    const frontmatterMatch = content.match(/^---\n([\s\S]*?)\n---\n([\s\S]*)$/);
+    if (!frontmatterMatch) {
+        // No frontmatter - treat entire content as memory
+        return {
+            name: basename(filePath, '.md'),
+            type: 'general',
+            content: content.trim(),
+            source: 'openclaw',
+        };
+    }
+    const [, yaml, body] = frontmatterMatch;
+    const metadata = {};
+    for (const line of yaml.split('\n')) {
+        const colonIdx = line.indexOf(':');
+        if (colonIdx > 0) {
+            const key = line.slice(0, colonIdx).trim();
+            const value = line.slice(colonIdx + 1).trim();
+            metadata[key] = value;
+        }
+    }
+    return {
+        name: metadata['name'] ?? basename(filePath, '.md'),
+        type: metadata['type'] ?? 'general',
+        content: body.trim(),
+        description: metadata['description'],
+        source: 'openclaw',
+    };
+}
+/**
+ * Import memories from OpenClaw.
+ */
+async function importOpenClawMemories(basePath) {
+    const memories = [];
+    const errors = [];
+    // Check for MEMORY.md index
+    const indexPath = join(basePath, 'MEMORY.md');
+    let memoryPaths = [];
+    try {
+        const indexContent = await readFile(indexPath, 'utf8');
+        memoryPaths = parseMemoryIndex(indexContent, basePath);
+    }
+    catch {
+        // No index - scan memory/ directory
+        const memoryDir = join(basePath, 'memory');
+        try {
+            const files = await readdir(memoryDir);
+            memoryPaths = files
+                .filter(f => f.endsWith('.md'))
+                .map(f => join(memoryDir, f));
+        }
+        catch {
+            errors.push('No MEMORY.md index or memory/ directory found');
+            return { memories, errors };
+        }
+    }
+    // Parse each memory file
+    for (const memPath of memoryPaths) {
+        try {
+            const content = await readFile(memPath, 'utf8');
+            const parsed = parseOpenClawMemory(content, memPath);
+            if (parsed)
+                memories.push(parsed);
+        }
+        catch (err) {
+            errors.push(`Failed to read ${memPath}: ${err instanceof Error ? err.message : String(err)}`);
+        }
+    }
+    return { memories, errors };
+}
+/**
+ * Discover OpenClaw skills (SKILL.md files).
+ */
+async function importOpenClawSkills(basePath) {
+    const skills = [];
+    const errors = [];
+    const skillDirs = [
+        join(basePath, 'skills'),
+        join(basePath, 'agent-skills'),
+    ];
+    for (const dir of skillDirs) {
+        try {
+            const entries = await readdir(dir, { withFileTypes: true });
+            for (const entry of entries) {
+                if (entry.isDirectory()) {
+                    const skillFile = join(dir, entry.name, 'SKILL.md');
+                    try {
+                        await stat(skillFile);
+                        skills.push(entry.name);
+                    }
+                    catch {
+                        // No SKILL.md in this dir
+                    }
+                }
+            }
+        }
+        catch {
+            // Directory doesn't exist
+        }
+    }
+    return { skills, errors };
+}
+/**
+ * Parse a ChatGPT conversation export (JSON format).
+ */
+function parseChatGPTExport(content) {
+    const conversations = [];
+    const errors = [];
+    try {
+        const data = JSON.parse(content);
+        if (!Array.isArray(data)) {
+            errors.push('Expected array of conversations in ChatGPT export');
+            return { conversations, errors };
+        }
+        for (const conv of data) {
+            try {
+                const messages = [];
+                if (conv.mapping) {
+                    // Extract messages from mapping (ChatGPT's internal format)
+                    const sorted = Object.values(conv.mapping)
+                        .filter(node => node.message?.content?.parts?.length)
+                        .sort((a, b) => (a.message?.create_time ?? 0) - (b.message?.create_time ?? 0));
+                    for (const node of sorted) {
+                        const msg = node.message;
+                        if (!msg?.author?.role || !msg.content?.parts?.length)
+                            continue;
+                        const role = msg.author.role === 'user' ? 'user'
+                            : msg.author.role === 'assistant' ? 'assistant'
+                                : 'system';
+                        messages.push({
+                            role,
+                            content: msg.content.parts.join('\n'),
+                            timestamp: msg.create_time
+                                ? new Date(msg.create_time * 1000).toISOString()
+                                : undefined,
+                        });
+                    }
+                }
+                if (messages.length > 0) {
+                    conversations.push({
+                        id: `chatgpt-${conversations.length}`,
+                        title: conv.title ?? `Conversation ${conversations.length + 1}`,
+                        messages,
+                        source: 'chatgpt',
+                    });
+                }
+            }
+            catch (err) {
+                errors.push(`Failed to parse conversation: ${err instanceof Error ? err.message : String(err)}`);
+            }
+        }
+    }
+    catch (err) {
+        errors.push(`Invalid JSON: ${err instanceof Error ? err.message : String(err)}`);
+    }
+    return { conversations, errors };
+}
+/**
+ * Extract memories from ChatGPT conversations.
+ * Looks for patterns: preferences, decisions, facts.
+ */
+function extractMemoriesFromConversations(conversations) {
+    const memories = [];
+    const seenTopics = new Set();
+    for (const conv of conversations) {
+        const userMessages = conv.messages.filter(m => m.role === 'user');
+        // Extract preferences
+        for (const msg of userMessages) {
+            const prefMatch = msg.content.match(/\b(?:I (?:use|prefer|like|work with|always))\s+(\w[\w\s.-]{2,30})/i);
+            if (prefMatch && !seenTopics.has(prefMatch[1].toLowerCase())) {
+                seenTopics.add(prefMatch[1].toLowerCase());
+                memories.push({
+                    name: `preference-${prefMatch[1].toLowerCase().replace(/\s+/g, '-')}`,
+                    type: 'user',
+                    content: `User prefers: ${prefMatch[1]}. Context: "${msg.content.slice(0, 100)}"`,
+                    source: 'chatgpt',
+                });
+            }
+        }
+        // Extract topic as a high-level memory if conversation is substantial
+        if (userMessages.length >= 3 && conv.title) {
+            const topicKey = conv.title.toLowerCase().slice(0, 30);
+            if (!seenTopics.has(topicKey)) {
+                seenTopics.add(topicKey);
+                memories.push({
+                    name: `topic-${topicKey.replace(/[^\w]/g, '-')}`,
+                    type: 'project',
+                    content: `Discussion about: ${conv.title}. ${userMessages.length} messages. First question: "${userMessages[0].content.slice(0, 100)}"`,
+                    source: 'chatgpt',
+                });
+            }
+        }
+    }
+    return memories;
+}
+// Aider Import
+/**
+ * Parse Aider chat history files (.aider.chat.history.md).
+ */
+async function importAiderHistory(basePath) {
+    const conversations = [];
+    const errors = [];
+    const historyFiles = [
+        join(basePath, '.aider.chat.history.md'),
+        join(basePath, '.aider', 'chat-history.md'),
+    ];
+    for (const histFile of historyFiles) {
+        try {
+            const content = await readFile(histFile, 'utf8');
+            const sections = content.split(/^#{1,2}\s+/m).filter(Boolean);
+            for (const section of sections) {
+                const lines = section.split('\n');
+                const title = lines[0]?.trim() ?? 'Aider session';
+                const messages = [];
+                let currentRole = null;
+                let currentContent = [];
+                for (const line of lines.slice(1)) {
+                    if (line.startsWith('> ') || line.startsWith('#### ')) {
+                        // Save previous message
+                        if (currentRole && currentContent.length > 0) {
+                            messages.push({ role: currentRole, content: currentContent.join('\n').trim() });
+                            currentContent = [];
+                        }
+                        currentRole = line.startsWith('> ') ? 'user' : 'assistant';
+                        currentContent.push(line.replace(/^> |^#### /, ''));
+                    }
+                    else if (currentRole) {
+                        currentContent.push(line);
+                    }
+                }
+                // Save last message
+                if (currentRole && currentContent.length > 0) {
+                    messages.push({ role: currentRole, content: currentContent.join('\n').trim() });
+                }
+                if (messages.length > 0) {
+                    conversations.push({
+                        id: `aider-${conversations.length}`,
+                        title,
+                        messages,
+                        source: 'aider',
+                    });
+                }
+            }
+        }
+        catch {
+            // File doesn't exist - try next
+        }
+    }
+    if (conversations.length === 0) {
+        errors.push('No Aider chat history found');
+    }
+    return { conversations, errors };
+}
+// Main Import Function
+/**
+ * Import data from another AI assistant into profClaw.
+ * Returns a summary of what was imported.
+ */
+export async function importFrom(source, options = {}) {
+    const start = Date.now();
+    const result = {
+        source,
+        memories: 0,
+        conversations: 0,
+        skills: 0,
+        preferences: 0,
+        errors: [],
+        duration: 0,
+    };
+    log.info(`Starting import from ${source}`, { path: options.path, file: options.file, dryRun: options.dryRun });
+    try {
+        switch (source) {
+            case 'openclaw': {
+                const basePath = findOpenClawPath(options.path);
+                // Import memories
+                const { memories, errors: memErrors } = await importOpenClawMemories(basePath);
+                result.errors.push(...memErrors);
+                if (!options.dryRun) {
+                    // Persist memories to experience store
+                    const { recordExperience } = await import('../../memory/experience-store.js');
+                    for (const mem of memories) {
+                        try {
+                            await recordExperience({
+                                type: mem.type === 'user' ? 'user_preference'
+                                    : mem.type === 'feedback' ? 'task_solution'
+                                        : 'task_solution',
+                                intent: mem.name,
+                                solution: { content: mem.content, importedFrom: 'openclaw' },
+                                successScore: 0.8,
+                                tags: ['imported', 'openclaw', mem.type],
+                                sourceConversationId: 'import-openclaw',
+                                userId: undefined,
+                            });
+                            result.memories++;
+                        }
+                        catch (err) {
+                            result.errors.push(`Failed to persist memory ${mem.name}: ${err instanceof Error ? err.message : String(err)}`);
+                        }
+                    }
+                }
+                else {
+                    result.memories = memories.length;
+                }
+                // Import skills
+                const { skills, errors: skillErrors } = await importOpenClawSkills(basePath);
+                result.errors.push(...skillErrors);
+                result.skills = skills.length;
+                // Extract preferences from memories
+                const prefMemories = memories.filter(m => m.type === 'user' || m.type === 'feedback');
+                result.preferences = prefMemories.length;
+                break;
+            }
+            case 'chatgpt': {
+                const filePath = options.file ?? options.path;
+                if (!filePath) {
+                    result.errors.push('ChatGPT import requires --file <path> pointing to conversations.json export');
+                    break;
+                }
+                const content = await readFile(filePath, 'utf8');
+                const { conversations, errors } = parseChatGPTExport(content);
+                result.errors.push(...errors);
+                result.conversations = conversations.length;
+                // Extract memories from conversations
+                const extracted = extractMemoriesFromConversations(conversations);
+                if (!options.dryRun) {
+                    const { recordExperience } = await import('../../memory/experience-store.js');
+                    for (const mem of extracted) {
+                        try {
+                            await recordExperience({
+                                type: mem.type === 'user' ? 'user_preference' : 'task_solution',
+                                intent: mem.name,
+                                solution: { content: mem.content, importedFrom: 'chatgpt' },
+                                successScore: 0.7,
+                                tags: ['imported', 'chatgpt', mem.type],
+                                sourceConversationId: 'import-chatgpt',
+                            });
+                            result.memories++;
+                        }
+                        catch {
+                            // Skip individual failures
+                        }
+                    }
+                }
+                else {
+                    result.memories = extracted.length;
+                }
+                result.preferences = extracted.filter(m => m.type === 'user').length;
+                break;
+            }
+            case 'aider': {
+                const basePath = options.path ?? process.cwd();
+                const { conversations, errors } = await importAiderHistory(basePath);
+                result.errors.push(...errors);
+                result.conversations = conversations.length;
+                // Extract solution patterns from Aider conversations
+                const memories = extractMemoriesFromConversations(conversations);
+                if (!options.dryRun) {
+                    const { recordExperience } = await import('../../memory/experience-store.js');
+                    for (const mem of memories) {
+                        try {
+                            await recordExperience({
+                                type: 'task_solution',
+                                intent: mem.name,
+                                solution: { content: mem.content, importedFrom: 'aider' },
+                                successScore: 0.7,
+                                tags: ['imported', 'aider'],
+                                sourceConversationId: 'import-aider',
+                            });
+                            result.memories++;
+                        }
+                        catch {
+                            // Skip individual failures
+                        }
+                    }
+                }
+                else {
+                    result.memories = memories.length;
+                }
+                break;
+            }
+        }
+    }
+    catch (error) {
+        result.errors.push(`Import failed: ${error instanceof Error ? error.message : String(error)}`);
+    }
+    result.duration = Date.now() - start;
+    log.info(`Import from ${source} complete`, {
+        memories: result.memories,
+        conversations: result.conversations,
+        skills: result.skills,
+        preferences: result.preferences,
+        errors: result.errors.length,
+        duration: result.duration,
+    });
+    return result;
+}
+/**
+ * List available import sources and their expected paths.
+ */
+export function listImportSources() {
+    return [
+        {
+            source: 'openclaw',
+            description: 'Import memories, skills, and preferences from OpenClaw',
+            defaultPath: join(homedir(), '.openclaw'),
+            format: 'MEMORY.md + memory/*.md files',
+        },
+        {
+            source: 'chatgpt',
+            description: 'Import conversations and extract preferences from ChatGPT export',
+            defaultPath: 'conversations.json (from Settings > Data Controls > Export Data)',
+            format: 'JSON array of conversations',
+        },
+        {
+            source: 'aider',
+            description: 'Import coding session history from Aider',
+            defaultPath: '.aider.chat.history.md in project root',
+            format: 'Markdown chat history',
+        },
+    ];
+}
+// CLI Command
+export function importCommands() {
+    const cmd = new Command('import')
+        .description('Import data from other AI assistants (OpenClaw, ChatGPT, Aider)');
+    cmd
+        .command('run')
+        .description('Import data from a source')
+        .requiredOption('--from <source>', 'Source to import from: openclaw, chatgpt, aider')
+        .option('--path <path>', 'Path to source data directory')
+        .option('--file <file>', 'Path to export file (for chatgpt)')
+        .option('--dry-run', 'Preview what would be imported without persisting', false)
+        .action(async (options) => {
+        const source = options.from;
+        if (!['openclaw', 'chatgpt', 'aider'].includes(source)) {
+            console.error(`Unknown source: ${source}. Use: openclaw, chatgpt, aider`);
+            process.exit(1);
+        }
+        console.log(`Importing from ${source}${options.dryRun ? ' (dry run)' : ''}...`);
+        const result = await importFrom(source, {
+            path: options.path,
+            file: options.file,
+            dryRun: options.dryRun,
+        });
+        console.log(`\nImport ${result.errors.length === 0 ? 'complete' : 'finished with errors'}:`);
+        console.log(`  Memories:      ${result.memories}`);
+        console.log(`  Conversations: ${result.conversations}`);
+        console.log(`  Skills:        ${result.skills}`);
+        console.log(`  Preferences:   ${result.preferences}`);
+        console.log(`  Duration:      ${result.duration}ms`);
+        if (result.errors.length > 0) {
+            console.log(`\n  Errors (${result.errors.length}):`);
+            for (const err of result.errors.slice(0, 5)) {
+                console.log(`    - ${err}`);
+            }
+            if (result.errors.length > 5) {
+                console.log(`    ... and ${result.errors.length - 5} more`);
+            }
+        }
+    });
+    cmd
+        .command('sources')
+        .description('List available import sources')
+        .action(() => {
+        const sources = listImportSources();
+        console.log('Available import sources:\n');
+        for (const s of sources) {
+            console.log(`  ${s.source}`);
+            console.log(`    ${s.description}`);
+            console.log(`    Default path: ${s.defaultPath}`);
+            console.log(`    Format: ${s.format}\n`);
+        }
+    });
+    return cmd;
+}
+//# sourceMappingURL=import.js.map

package/dist/cli/commands/init.d.ts

@@ -0,0 +1,13 @@
+/**
+ * CLI Init Command
+ *
+ * Scans the current project directory and generates a `.profclaw/context.md`
+ * file describing the tech stack, structure, build/test commands, and conventions.
+ *
+ * Usage:
+ *   profclaw init          # interactive (asks before overwriting)
+ *   profclaw init --force  # overwrite without prompting
+ */
+import { Command } from 'commander';
+export declare function initCommand(): Command;
+//# sourceMappingURL=init.d.ts.map