lexiforest/curl_cffi/imp-fetch

curl-cffi CLI — The Impersonated Web Fetch Skill

Use the curl-cffi CLI as a replacement for the web_fetch tool. It fetches web pages with browser impersonation, bypassing TLS fingerprint-based blocking. It supports HTTP/2 and HTTP/3 out of the box.

When to use

• When web_fetch is unavailable or returns blocked/empty responses
• When a website requires a real browser TLS fingerprint
• When you need to fetch API endpoints, HTML pages, or JSON resources
• When you need to replay .http or .har files in batch

Usage

curl-cffi METHOD URL [REQUEST_ITEMS...] [FLAGS]

• METHOD is required: get, post, put, delete, patch, head, options, trace, query (case-insensitive).
• URL is required. Bare domains default to https://. A leading colon is a localhost shortcut (:3000 → http://localhost:3000). An explicit port other than 443 uses http://.

Basic usage

# Simple GET (impersonates Chrome by default)
curl-cffi get https://httpbin.org/get
 
# POST JSON data (`:=` for non-string values)
curl-cffi post https://httpbin.org/post name=John age:=30
 
# POST form data
curl-cffi post --form https://httpbin.org/post name=John
 
# Custom header
curl-cffi get https://httpbin.org/get X-My-Header:value
 
# Impersonate Safari instead of Chrome
curl-cffi get -i safari https://tls.browserleaks.com/json
 
# HTTP/3
curl-cffi get --http3 https://fp.impersonate.pro/api/http3
 
# Localhost shortcut
curl-cffi get :8000/api/health

Output control

When connected to a terminal, default output includes response headers and body with syntax highlighting. When piped, only the body is printed as plain text.

# Body only (good for piping/parsing)
curl-cffi get --body https://httpbin.org/get
 
# Headers only
curl-cffi get --headers https://httpbin.org/get
 
# Full verbose output (request + response headers and body)
curl-cffi post -v https://httpbin.org/post name=test
 
# Fine-grained: request headers + response headers only
curl-cffi get -p Hh https://httpbin.org/get

Flags reference

Request item syntax

Batch execution

The run subcommand executes multiple requests from a file:

# Run requests from an .http file (shared session by default)
curl-cffi run requests.http
 
# Replay a HAR file exported from Chrome DevTools
curl-cffi run session.har
 
# Independent requests (no shared cookies/connections)
curl-cffi run --no-session requests.http

Supported formats: .http / .rest (HTTP Request in Editor format) and .har (HTTP Archive).

Diagnostics

curl-cffi doctor

Prints Python version, platform, curl_cffi version, and libcurl version.