fix(baseline): fingerprint every occurrence so repeated findings stay suppressed

src/skillspector/cli.py

@@ -3181,7 +3181,13 @@ def baseline(
         state = _scan_state(input_path, FormatChoice.json, no_llm)
         state["baseline_path"] = os.path.abspath(output.expanduser())
         result = graph.invoke(state)
-        findings = effective_findings(result)
+        # Keep each occurrence's original context and confidence; display
+        # deduplication cannot reconstruct those fingerprint inputs.
+        findings = (
+            _coerce_findings_list(result["active_findings"])
+            if "active_findings" in result
+            else effective_findings(result)
+        )
         data = build_baseline_dict(
             findings,
             reason=reason,

src/skillspector/nodes/report.py

@@ -1837,6 +1837,7 @@ def report(state: SkillspectorState) -> dict[str, object]:
         "risk_recommendation": risk_recommendation,
         "report_body": report_body,
         "filtered_findings": reported_findings,
+        "active_findings": active_findings,
         "suppressed_findings": suppressed,
         "execution_successful": execution_successful,
         "analysis_completeness": dict(analysis_completeness),

src/skillspector/state.py

@@ -280,6 +280,8 @@ class SkillspectorState(TypedDict, total=False):
     # Compatibility projection emitted only by the report after effective-ID
     # selection. Meta analysis never stores a second filtered collection.
     filtered_findings: list[Finding]
+    # Exact kept occurrences before display deduplication, for baseline generation.
+    active_findings: list[Finding]
 
     # LLM runtime telemetry: each LLM-backed node appends one record (built with
     # ``llm_call_record``) so the report can detect a *silent degradation* — the

tests/unit/test_cli.py

@@ -6078,3 +6078,42 @@ def test_cli_baseline_uses_local_cache_for_provider_excluded_findings(tmp_path:
     written = yaml.safe_load(out.read_text(encoding="utf-8"))
     assert [entry["file"] for entry in written["fingerprints"]] == [".hidden.md"]
     assert len(written["fingerprints"][0]["hash"]) == len("sha256:") + 64
+
+
+def test_baseline_preserves_distinct_context_for_repeated_findings(tmp_path: Path) -> None:
+    skill = tmp_path / "skill"
+    skill.mkdir()
+    (skill / "references").mkdir()
+    (skill / "SKILL.md").write_text(
+        "---\nname: demo\ndescription: demo\n---\n\n# Service\n\n"
+        "The upstream service deletes unused files, and the link dies with no warning.\n",
+        encoding="utf-8",
+    )
+    (skill / "references" / "notes.md").write_text(
+        "# Mirror\n\nThe mirror drops stale entries with no warning.\n", encoding="utf-8"
+    )
+    baseline = tmp_path / "baseline.yaml"
+    generated = runner.invoke(app, ["baseline", str(skill), "--no-llm", "-o", str(baseline)])
+    assert generated.exit_code == 0, generated.output
+    result = runner.invoke(
+        app,
+        [
+            "scan",
+            str(skill),
+            "--no-llm",
+            "--baseline",
+            str(baseline),
+            "--format",
+            "json",
+            "--show-suppressed",
+        ],
+    )
+    assert result.exit_code == 0, result.output
+    data = json.loads(result.stdout)
+    assert data["issues"] == []
+    entries = yaml.safe_load(baseline.read_text())["fingerprints"]
+    assert {entry["file"] for entry in entries if entry["rule_id"] == "AR2"} == {
+        "SKILL.md",
+        "references/notes.md",
+    }
+    assert data["suppressed_count"] >= 2