AWS SigV4 Request Builder

Generate Python code to call undocumented AWS APIs using AWS Signature Version 4 authentication from cURL requests.

When to Use

Use this skill when users need to:

Call undocumented AWS APIs from Python
Convert browser network requests to authenticated Python code
Build API clients for AWS services without official SDK support

How to Use

Step 1: Get cURL from User

Ask the user for the cURL command (from browser dev tools: Network tab → Right-click → Copy as cURL).

Step 2: Extract Information

From the cURL command, extract:

URL: Full endpoint URL
Region: Extract from URL (e.g., us-east-1 from service.us-east-1.amazonaws.com)
Service name: Extract from authorization header SignedHeaders or URL (e.g., /us-east-1/q/aws4_request → "q")
Operation name: Extract from x-amz-target header (part after the dot, e.g., CreateAssignment from AmazonQDeveloperService.CreateAssignment)
Function name: Convert operation name to snake_case (e.g., CreateAssignment → create_assignment)
HTTP method: From -X flag or default to POST if --data present
Headers: Content-Type and x-amz-target (exclude authorization, cookie, user-agent, sec-*, origin, referer, priority)
Payload: From --data-raw or --data
Parameters: Extract keys from JSON payload

Step 3: Generate Python Code

Create a Python function following this pattern:

python

import boto3
import json
from botocore.awsrequest import AWSRequest
from botocore.auth import SigV4Auth
import requests


def function_name(param1, param2):
    """Call AWS API with description"""

    session = boto3.Session()
    credentials = session.get_credentials()

    payload = {
        "key1": param1,
        "key2": param2
    }

    request = AWSRequest(
        method="POST",
        url="https://service.region.amazonaws.com/",
        data=json.dumps(payload),
        headers={
            "Content-Type": "application/x-amz-json-1.0",
            "x-amz-target": "ServiceName.OperationName",
        },
    )

    SigV4Auth(credentials, "service-name", "region").add_auth(request)

    response = requests.post(request.url, headers=dict(request.headers), data=request.body)

    return response

Example

User provides cURL:

bash

curl 'https://codewhisperer.us-east-1.amazonaws.com/' \
  -H 'authorization: AWS4-HMAC-SHA256 Credential=.../us-east-1/q/aws4_request, ...' \
  -H 'content-type: application/x-amz-json-1.0' \
  -H 'x-amz-target: AmazonQDeveloperService.CreateAssignment' \
  --data-raw '{"principalId":"123","principalType":"USER","subscriptionType":"Q_DEVELOPER_STANDALONE_PRO"}'

Auto-extracted:

Service: "q" (from authorization header: /us-east-1/q/aws4_request)
Region: "us-east-1" (from URL)
Operation: "CreateAssignment" (from x-amz-target)
Function name: "create_assignment" (snake_case of operation)
Parameters: principalId, principalType, subscriptionType (from payload)

Generated code:

python

import boto3
import json
from botocore.awsrequest import AWSRequest
from botocore.auth import SigV4Auth
import requests


def create_assignment(principal_id, principal_type="USER", subscription_type="Q_DEVELOPER_STANDALONE_PRO"):
    """Create Q Developer subscription assignment"""

    session = boto3.Session()
    credentials = session.get_credentials()

    payload = {
        "principalId": principal_id,
        "principalType": principal_type,
        "subscriptionType": subscription_type
    }

    request = AWSRequest(
        method="POST",
        url="https://codewhisperer.us-east-1.amazonaws.com/",
        data=json.dumps(payload),
        headers={
            "Content-Type": "application/x-amz-json-1.0",
            "x-amz-target": "AmazonQDeveloperService.CreateAssignment",
        },
    )

    SigV4Auth(credentials, "q", "us-east-1").add_auth(request)

    response = requests.post(request.url, headers=dict(request.headers), data=request.body)

    return response

Prerequisites

Generated code requires:

Python 3.6+
boto3: pip install boto3
requests: pip install requests
AWS credentials configured

Search AI Tools

aws-sigv4-req-builder

Install this agent skill to your Project

SKILL.md