AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary
Python ai-promptsai-skillblueskyclaudeclaude-codeclawhubdeep-researchhackernewsinstagramopenclawpolymarketrecencyredditresearchsocial-mediatiktoktrendstwitterweb-searchyoutube
"""Supplemental X lanes must stay inside the effective research plan."""
from unittest.mock import patch
from uuid import uuid4
import pytest
from lib import dates, pipeline, x_envelope
@pytest.mark.parametrize("depth", ["default", "deep"])
@pytest.mark.parametrize("lane", ["backend", "envelope"])
@pytest.mark.parametrize(
"requested,excluded,planned,allowed",
[
pytest.param(["reddit"], "x", ["reddit"], False, id="reddit-and-exclusion"),
pytest.param(["reddit"], "", ["reddit", "x"], False, id="requested-reddit"),
pytest.param(None, "x", ["reddit", "x"], False, id="excluded-x"),
pytest.param(["reddit", "x"], "x", ["reddit", "x"], False, id="exclusion-wins"),
pytest.param(["reddit", "x"], "", ["reddit"], False, id="plan-omits-x"),
pytest.param(None, "", ["reddit", "x"], True, id="available-x"),
pytest.param(["reddit", "x"], "", ["reddit", "x"], True, id="requested-x"),
pytest.param(["xquik"], "", ["reddit", "x"], True, id="requested-alias"),
],
)
def test_supplemental_lanes_obey_effective_sources(depth, lane, requested, excluded, planned, allowed):
topic = "widget development"
marker = uuid4().hex
window = dates.get_date_range(30)
posts = [
{
"id": str(index),
"url": f"https://x.com/example/status/{index}",
"author_handle": "example",
"text": f"{topic} {marker}: {label}",
"date": window[1],
"engagement": {"likes": 120, "retweets": 10, "replies": 5},
"relevance": 1.0,
}
for index, label in enumerate([
"shipping a compiler release with faster incremental builds and reduced memory consumption",
"community feedback suggests clearer tutorials explaining deployment authentication errors for beginners",
], start=1)
]
envelope = None
if lane == "envelope":
envelope = x_envelope.Envelope(
path="fixture.json", sha256="fixture", status="ok", error_category="",
provider="x-connector", window=window, topic_items=[],
lane_calls=[
x_envelope.EnvelopeCall(0, "from", ["example"], [posts[0]]),
x_envelope.EnvelopeCall(1, "mention", ["example"], [posts[1]]),
],
counters={}, accepted=2, total=2,
lane_counts={"from": 1, "mention": 1}, call_lanes=["from", "mention"],
)
config = {
"X_BEARER_TOKEN": "dummy-bearer",
"LAST30DAYS_X_BACKEND": "xapi",
"EXCLUDE_SOURCES": excluded,
}
plan = {
"intent": "opinion", "freshness_mode": "balanced_recent", "cluster_mode": "debate",
"subqueries": [{
"label": "primary", "search_query": topic,
"ranking_query": topic, "sources": planned,
}],
}
with (
patch.object(pipeline.reddit_public, "search_reddit_public", autospec=True, return_value=[]),
patch.object(pipeline.x_api, "search_x", autospec=True, return_value={"items": []}),
patch.object(pipeline.x_api, "search_handles", autospec=True, return_value=[posts[0]]) as from_lane,
patch.object(pipeline.x_api, "search_mentions", autospec=True, return_value=[posts[1]]) as about_lane,
):
report = pipeline.run(
topic=topic, config=config, depth=depth, requested_sources=requested,
external_plan=plan, x_handle="example", x_posts=envelope,
web_backend="none", save_dir="",
)
evidence = report.items_by_source.get("x", [])
if allowed:
assert {item.url for item in evidence} == {post["url"] for post in posts}
assert all(marker in item.body for item in evidence)
else:
assert not evidence
if allowed and lane == "backend":
from_lane.assert_called_once()
about_lane.assert_called_once()
else:
from_lane.assert_not_called()
about_lane.assert_not_called()
if envelope is not None:
remaining = envelope.take_lanes()
if allowed:
assert remaining is None
else:
assert remaining == envelope.lane_calls