ghostfetchlisted
Install: claude install-skill iArsalanshah/GhostFetch
# GhostFetch Skill
Fetch web content from sites that block AI agents. Uses a stealthy headless browser with advanced fingerprinting to bypass anti-bot protections and returns clean Markdown.
## When to Use
- Fetching content from X.com/Twitter posts
- Reading articles from sites that block bots
- Extracting content from JavaScript-heavy sites
- Getting clean Markdown from any webpage for LLM consumption
## Prerequisites
GhostFetch must be running as a service. Start it with:
```bash
# Option 1: If installed via pip
ghostfetch serve
# Option 2: Docker
docker run -p 8000:8000 iarsalanshah/ghostfetch
```
## Usage
### Synchronous Fetch (Recommended)
Use the `/fetch/sync` endpoint for simple, blocking requests:
```bash
curl "http://localhost:8000/fetch/sync?url=https://example.com"
```
### Python
```python
import requests
def ghostfetch(url: str, timeout: float = 120.0) -> dict:
"""
Fetch content from a URL using GhostFetch.
Returns:
dict with 'metadata' and 'markdown' keys
"""
response = requests.post(
"http://localhost:8000/fetch/sync",
json={"url": url, "timeout": timeout}
)
response.raise_for_status()
return response.json()
# Example
result = ghostfetch("https://x.com/user/status/123")
print(result["markdown"])
```
### With SDK
```python
from ghostfetch import fetch
result = fetch("https://x.com/user/status/123")
print(result["metadata"]["title"])
print(result["markdown"])
```
## Response Format
`