Skip to content

AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary

fix: preserve slash topics outside comparison mode (#1211)

Ilia Alshanetsky4e9ac0ebf2b2
diff --git a/changelog.d/+cr-006-comparison-intent.fixed.md b/changelog.d/+cr-006-comparison-intent.fixed.md
new file mode 100644
index 00000000..a24777c6
--- /dev/null
+++ b/changelog.d/+cr-006-comparison-intent.fixed.md
@@ -0,0 +1 @@
+Keep slash-containing research topics such as `CI/CD`, repository names, and URLs intact instead of launching unintended competitor comparisons. Preserve supported comparison forms, including `React/Vue/Svelte`, and slashes inside explicitly compared entities.
diff --git a/skills/last30days/scripts/lib/planner.py b/skills/last30days/scripts/lib/planner.py
index f0b1a703..08d9d7fe 100644
--- a/skills/last30days/scripts/lib/planner.py
+++ b/skills/last30days/scripts/lib/planner.py
@@ -759,12 +759,15 @@ def _fallback_plan(
)
+_SLASH_COMPARISON = re.compile(r"\b[A-Z][a-z]{2,}(?:/[A-Z][a-z]{2,})+\b")
+
+
def _infer_intent(topic: str) -> str:
text = topic.lower().strip()
if re.search(r"\b(vs|versus|compare|compared to|difference between)\b", text):
return "comparison"
# Slash-separated proper nouns: "React/Vue/Svelte" (not URLs, not acronyms like CI/CD or I/O)
- if not re.search(r"https?://", topic) and re.search(r"\b[A-Z][a-z]{2,}(?:/[A-Z][a-z]{2,})+\b", topic):
+ if not re.search(r"https?://", topic) and _SLASH_COMPARISON.search(topic):
return "comparison"
if re.search(r"\b(odds|predict|prediction|forecast|chance|probability|will .* win)\b", text):
return "prediction"
@@ -901,7 +904,13 @@ def _comparison_entities(topic: str, *, uncapped: bool = False) -> list[str]:
Caps at ``competitors.COMPARISON_ENTITY_MAX`` unless ``uncapped`` (caller
truncates and may warn about dropped entities).
+
+ Standalone comparator tokens are syntax, including repeated tokens.
+ Compact ``vs.`` separates entities only after a nonempty left entity.
"""
+ if _infer_intent(topic) != "comparison":
+ return []
+
# "difference between X and Y" -> "X vs Y" (replace "and" only in this context)
normalized = re.sub(
r"\bdifference between\s+(.+?)\s+and\s+",
@@ -910,11 +919,25 @@ def _comparison_entities(topic: str, *, uncapped: bool = False) -> list[str]:
flags=re.I,
)
normalized = re.sub(r"\b(compared to)\b", " vs ", normalized, flags=re.I)
- parts = [
- part.strip(" \t\r\n?.,:;!()[]{}\"'")
- for part in re.split(r"\bvs\.?\b|\bversus\b|/", normalized, flags=re.I)
- if part.strip(" \t\r\n?.,:;!()[]{}\"'")
- ]
+ separator = r"(?<!\S)(?:(?P<standalone>vs\.?|versus)(?!\S)|vs\.(?=\S))"
+ if not re.search(separator, normalized, flags=re.I):
+ if re.search(r"https?://", normalized):
+ return []
+ normalized = _SLASH_COMPARISON.sub(
+ lambda match: match.group(0).replace("/", " vs "), normalized,
+ )
+ trim = " \t\r\n?.,:;!()[]{}\"'"
+ parts = []
+ part_start = 0
+ for match in re.finditer(separator, normalized, flags=re.I):
+ part = normalized[part_start:match.start()].strip(trim)
+ if part:
+ parts.append(part)
+ if part or match.group("standalone") is not None:
+ part_start = match.end()
+ last_part = normalized[part_start:].strip(trim)
+ if last_part:
+ parts.append(last_part)
# Strip trailing context from parts ("Svelte for frontend in 2026" -> "Svelte")
if len(parts) < 2:
return []
diff --git a/tests/test_comparison_intent.py b/tests/test_comparison_intent.py
new file mode 100644
index 00000000..b9bb4906
--- /dev/null
+++ b/tests/test_comparison_intent.py
@@ -0,0 +1,127 @@
+import json
+import sys
+
+import pytest
+
+import last30days as cli
+from lib import planner
+
+
+SINGLE_TOPICS = [
+ "CI/CD",
+ "I/O performance",
+ "openai/openai-python",
+ "https://example.com",
+ "https://example.com/React/Vue",
+ "https://example.com/vs/topic",
+ "https://example.com/vs.Vue",
+ "React vsVue",
+ "vs.code",
+ "vs.code/extension",
+ "compare CI/CD workflows",
+]
+
+COMPARISONS = [
+ ("React/Vue/Svelte", ["React", "Vue", "Svelte"]),
+ ("React vs Vue", ["React", "Vue"]),
+ ("React vs. Vue", ["React", "Vue"]),
+ ("React vs. vs. Vue", ["React", "Vue"]),
+ ("React vs.Vue", ["React", "Vue"]),
+ ("React VS.Vue", ["React", "Vue"]),
+ ("React versus Vue", ["React", "Vue"]),
+ ("vs.code versus Vue", ["vs.code", "Vue"]),
+ ("vs.code vs.Vue", ["vs.code", "Vue"]),
+ ("vs.code/extension versus Vue", ["vs.code/extension", "Vue"]),
+ ("React versus vs.code", ["React", "vs.code"]),
+ ("React vs.vs.code", ["React", "vs.code"]),
+ ("React versus vs.code/extension", ["React", "vs.code/extension"]),
+ ("vs.code/extension versus vs.code/other", ["vs.code/extension", "vs.code/other"]),
+ ("https://example.com/vs.code versus owner/repo",
+ ["https://example.com/vs.code", "owner/repo"]),
+ ("React versus https://example.com/vs.code", ["React", "https://example.com/vs.code"]),
+ ("React compared to Vue", ["React", "Vue"]),
+ ("difference between React and Vue", ["React", "Vue"]),
+ ("React/Vue/Svelte for CI/CD", ["React", "Vue", "Svelte"]),
+ ("CI/CD vs I/O", ["CI/CD", "I/O"]),
+ ("CI/CD vs.I/O", ["CI/CD", "I/O"]),
+ ("openai/openai-python vs anthropics/anthropic-sdk-python",
+ ["openai/openai-python", "anthropics/anthropic-sdk-python"]),
+]
+
+
[email protected]("topic", SINGLE_TOPICS)
[email protected]("uncapped", [False, True])
+def test_slashes_without_comparison_intent_do_not_produce_entities(topic, uncapped):
+ assert planner._comparison_entities(topic, uncapped=uncapped) == []
+
+
[email protected]("topic,entities", COMPARISONS)
+def test_comparison_entities_preserve_slashes_inside_explicit_entities(topic, entities):
+ assert planner._comparison_entities(topic) == entities
+
+
+def run_cli(monkeypatch, tmp_path, capsys):
+ monkeypatch.setattr(cli.env, "CONFIG_DIR", tmp_path)
+ monkeypatch.setattr(cli.env, "get_config", lambda **kwargs: {})
+ monkeypatch.setenv("PATH", "")
+ monkeypatch.setenv("LAST30DAYS_STORE", "")
+
+ def run(topic):
+ monkeypatch.setattr(sys, "argv", [
+ "last30days", topic, "--mock", "--quick", "--search=reddit",
+ "--no-browser-cookies", "--emit=json", "--json-profile=raw",
+ "--save-dir", "",
+ ])
+ result = cli.main()
+ captured = capsys.readouterr()
+ assert result == 0, captured.err
+ return json.loads(captured.out)
+
+ return run
+
+
[email protected]("topic", SINGLE_TOPICS)
+def test_main_preserves_slash_topics_as_single_research_runs(topic, run_cli):
+ payload = run_cli(topic)
+ assert payload.get("comparison") is not True
+ assert payload["topic"] == topic
+ assert payload["query_plan"]["raw_topic"] == topic
+
+
[email protected]("topic,entities", COMPARISONS)
+def test_main_keeps_supported_comparison_forms(topic, entities, run_cli):
+ payload = run_cli(topic)
+ assert payload["comparison"] is True
+ assert payload["entities"] == entities
+ assert [entry["report"]["topic"] for entry in payload["reports"]] == entities
+
+
[email protected]("left,right", [
+ pytest.param("React", "Vue", id="plain"),
+ pytest.param("vs.code", "vs.editor", id="dotted"),
+ pytest.param("org/repo", "other/lib", id="repository"),
+ pytest.param("vs.code/extension", "vs.editor/tool", id="dotted-repository"),
+ pytest.param("https://example.com/vs.code", "https://example.org/vs.editor", id="url"),
+])
+ pytest.param("{left} vs {right}", id="spaced-vs"),
+ pytest.param("{left} vs. {right}", id="spaced-dotted-vs"),
+ pytest.param("{left} versus {right}", id="versus"),
+ pytest.param("{left} vs.{right}", id="compact-vs"),
+ pytest.param("{left} vs vs {right}", id="repeated-vs"),
+ pytest.param("{left} vs. vs. {right}", id="repeated-dotted-vs"),
+ pytest.param("{left} versus versus {right}", id="repeated-versus"),
+ pytest.param("{left} vs versus vs. {right}", id="mixed-separators"),
+ pytest.param("vs {left} vs {right}", id="leading-vs"),
+ pytest.param("vs. {left} vs.{right}", id="leading-dotted-vs"),
+ pytest.param("versus {left} versus {right}", id="leading-versus"),
+ pytest.param("vs vs. versus {left} versus vs. vs {right}", id="leading-and-repeated"),
+])
+def test_separator_tokens_preserve_entity_names(left, right, form, run_cli):
+ topic = form.format(left=left, right=right)
+ assert planner._comparison_entities(topic) == [left, right]
+ payload = run_cli(topic)
+ assert payload["comparison"] is True
+ assert payload["entities"] == [left, right]
+ assert [entry["report"]["topic"] for entry in payload["reports"]] == [left, right]
diff --git a/tests/test_registers.py b/tests/test_registers.py
index 6ebbde80..0ca53fd7 100644
--- a/tests/test_registers.py
+++ b/tests/test_registers.py
@@ -269,7 +269,7 @@ def test_registers_do_not_shape_comparison_output():
@pytest.mark.parametrize(
"topic",
[
- "alpha/beta",
+ "React/Vue",
"alpha compared to beta",
"difference between alpha and beta",
],
@@ -282,6 +282,17 @@ def test_registers_use_canonical_comparison_detection(topic):
).name == "default"
[email protected]("topic", ["CI/CD", "alpha/beta", "https://example.com"])
+def test_slash_topics_keep_single_topic_register_selection(topic):
+ args = cli.build_parser().parse_args([topic])
+
+ assert cli._audience_register_for_run(
+ args, {"LAST30DAYS_REGISTER": "creator"}, None
+ ).name == "creator"
+ with pytest.raises(ValueError, match="unknown audience register 'board'"):
+ cli._audience_register_for_run(args, {"LAST30DAYS_REGISTER": "board"}, None)
+
+
def test_registered_html_excludes_source_failure_diagnostics():
report = fixture_report()
report.source_status["x"] = schema.SourceOutcome(