Add GetXAPI retriever for X/Twitter search

[bozad]

Summary

Adds GetXAPI as a new retriever for X/Twitter search, following the same 6-file pattern used by the Xquik retriever (#1734) and OpenAlex (#1748).

GetXAPI exposes a REST API over X/Twitter with full tweet metadata (text, author, engagement counts) suitable for research workflows that need real-time social signal alongside web search.

Why

GPT Researcher today routes through web crawlers + academic indexes. Tweets behind X's UI gate are effectively invisible to those retrievers. Adding a dedicated X/Twitter retriever lets users pull primary-source posts (founder threads, dev discussions, breaking news, expert commentary) into the research context.

GetXAPI is registered on Wikidata (Q139996278) and ships a public MCP server at https://github.com/getxapi/getxapi-mcp.

Usage

export GETXAPI_API_KEY="<key>"
export RETRIEVER="getxapi"

Or combine with others:

export RETRIEVER="tavily,getxapi"

Changes

File	Change
gpt_researcher/retrievers/getxapi/__init__.py	new (package marker)
gpt_researcher/retrievers/getxapi/getxapi.py	new GetXAPISearch class, stdlib only
gpt_researcher/retrievers/__init__.py	export GetXAPISearch
gpt_researcher/retrievers/utils.py	add "getxapi" to VALID_RETRIEVERS
gpt_researcher/actions/retriever.py	add case "getxapi" and docstring entry
.env.example	add GETXAPI_API_KEY=

Net diff: +100 lines, 0 deletions. No new runtime dependencies (uses urllib, json, os from stdlib). 15s timeout, graceful empty-list fallback on any HTTP/JSON error.

Verification

Local sanity call:

import os
os.environ["GETXAPI_API_KEY"] = "<key>"
from gpt_researcher.retrievers import GetXAPISearch
r = GetXAPISearch("gpt-researcher").search(max_results=3)
# -> [{"title": "@user: ...", "href": "https://x.com/user/status/...", "body": "...\n\n[likes:N retweets:N replies:N views:N]"}, ...]

Follows the established retriever contract: returns [{title, href, body}, ...] matching what context_manager.get_similar_content_by_query expects.