x2o — X Bookmark Export + AI Classification
Export X (Twitter) bookmarks → AI classify → Obsidian knowledge vault. All local, no third-party servers.
What's New (v0.4.0)
- Standalone reclassify: pick existing output folder, reclassify with different AI
- Reads markdown content directly (no X link dependency)
- t.co short URLs automatically resolved to real URLs
- X Notes (long-form tweets) full content extraction
Usage
Run the CLI script via npx tsx:
npx tsx ~/.claude/skills/x2o/scripts/x2o.ts \
--cookie "<X cookie string>" \
--provider openai \
--api-key "sk-..." \
--output ~/x2o-output \
--limit 100
# 也可以换成其他 provider：
# --provider claude --api-key "sk-ant-..."
# --provider gemini --api-key "..."
# --provider ollama --model llama3.2 （本地 Ollama 不需要 api-key）
Re-classify existing bookmarks (skip fetching):
If you already have an output folder that contains bookmarks.json, you can re-run AI classification without fetching again:
npx tsx ~/.claude/skills/x2o/scripts/x2o.ts \
--input ~/x2o-output \
--provider openai \
--api-key "sk-..." \
--output ~/x2o-output
Rebuild from Markdown vault (only .md files) and re-classify:
npx tsx ~/.claude/skills/x2o/scripts/x2o.ts \
--md-dir ~/x2o-output \
--provider openai \
--api-key "sk-..." \
--output ~/x2o-output
Notes:
- Passing a directory to
--input/--reclassify/--md-dirwill first look for<dir>/bookmarks.json.- If not found, it will try to parse X URLs from
.mdfiles and reconstruct a minimal bookmarks list.
Parameters
| Flag | Required | Description |
|---|---|---|
--cookie | Yes* | X browser cookie (must contain ct0) |
--input | Yes* | Path to existing bookmarks JSON or an output folder containing bookmarks.json (alternative to cookie) |
--reclassify | No | Alias of --input |
--md-dir | No | Path to a Markdown vault folder (only .md files). If the folder has bookmarks.json, it will use that first. |
--provider | No | AI provider (default: openai): openai, claude, deepseek, gemini, ollama, groq, moonshot, qwen, zhipu, siliconflow, mistral, together, fireworks, xai, openrouter, cohere, deepinfra, perplexity |
--api-key | Yes** | API key for the chosen provider (not required for ollama, or when using --fetch-only) |
--output | No | Output directory (default: ~/x2o-output) |
--limit | No | Max bookmarks to fetch (default: 800) |
--model | No | Override default model for the provider |
--base-url | No | Override API base URL |
--fetch-only | No | Only fetch bookmarks, skip classification |
* One of --cookie or --input is required
** Not required for ollama (local)
Workflow for Claude
When user wants to export bookmarks:
- Ask for their X cookie (guide them: F12 → Network → find any request → copy Cookie header)
- Ask which AI provider they want to use + API key
- Run the script with appropriate parameters
- Report results: how many bookmarks fetched, categories found, files generated
- Optionally open the output in Obsidian:
open "obsidian://open?path=<output-dir>"
Supported AI Providers (18+)
OpenAI, Claude, DeepSeek, Gemini, Ollama, OpenRouter, Moonshot, Qwen, Zhipu GLM, SiliconFlow, Groq, Mistral, Together AI, Fireworks, Grok (xAI), Cohere, DeepInfra, Perplexity