diff --git a/agentrace/checks.py b/agentrace/checks.py
index 13c437f..a78f4ca 100644
--- a/agentrace/checks.py
+++ b/agentrace/checks.py
@@ -96,18 +96,20 @@ def check_unverified_claim(run: AgentRun) -> list[Finding]:
         r"\bcould not (?:independently )?verify\b",
         r"\bunverified\b",
     ]
-    hits = []
+    hits = 0
+    first_context = ""
     for p in hedges:
         for m in re.finditer(p, run.result, re.I):
-            hits.append(_context(run.result, m.start()))
-            break
+            if hits == 0:
+                first_context = _context(run.result, m.start())
+            hits += 1
     if hits:
         return [
             Finding(
                 "hedged_claim",
                 "low",
-                f"{len(hits)} hedged claim(s). Fine if the hedge survives downstream, a problem if it gets flattened into fact.",
-                hits[0],
+                f"{hits} hedged claim(s). Fine if the hedge survives downstream, a problem if it gets flattened into fact.",
+                first_context,
             )
         ]
     return []
diff --git a/tests/test_agentrace.py b/tests/test_agentrace.py
index 96e5e8d..12d26fd 100644
--- a/tests/test_agentrace.py
+++ b/tests/test_agentrace.py
@@ -179,6 +179,12 @@ def test_hedged_claim_is_caught_but_only_low():
     assert findings and findings[0].severity == "low"
 
 
+def test_hedged_claim_counts_every_occurrence():
+    r = _run(result="probably one. probably two. probably three. " + "x" * 200)
+    findings = [f for f in analyse(r) if f.check == "hedged_claim"]
+    assert findings and "3 hedged claim(s)" in findings[0].message
+
+
 def test_many_urls_without_verification_is_flagged():
     urls = " ".join(f"https://example{i}.com/careers" for i in range(8))
     assert "unverified_urls" in _codes(_run(result=urls + " " + "x" * 200))