diff --git a/graphify/detect.py b/graphify/detect.py
index 98a49518..2f846460 100644
--- a/graphify/detect.py
+++ b/graphify/detect.py
@@ -1042,13 +1042,19 @@ def detect(root: Path, *, follow_symlinks: bool | None = None, google_workspace:
                 # Prune noise dirs in-place so os.walk never descends into them.
                 # Dot dirs are allowed — users often want .github/, .claude/, etc.
                 # Framework caches (.next, .nuxt, …) are caught by _is_noise_dir.
-                # When negation patterns (!) exist, skip directory-level ignore
-                # pruning so negated files inside can still be reached.
-                has_negation = any(p.startswith("!") for _, p in ignore_patterns)
+                # Negations need no special-casing here: _is_ignored already applies
+                # last-match-wins (so `!dir/` un-ignores a directory and it won't be
+                # pruned) and the gitignore parent-exclusion rule (a `!` cannot rescue
+                # a file beneath an excluded dir), so descending an ignored directory to
+                # look for a re-included file is never necessary. The previous blanket
+                # `has_negation` disabled directory pruning for EVERY ignored dir whenever
+                # any `!` rule existed — e.g. a single `!docs/**` made the walk descend
+                # bin/, obj/, wwwroot/, generated/, … : a pathological slowdown on large
+                # repos for no correctness gain.
                 dirnames[:] = [
                     d for d in dirnames
                     if not _is_noise_dir(d, dp)
-                    and (has_negation or not _is_ignored(dp / d, root, ignore_patterns, _cache=ignore_cache))
+                    and not _is_ignored(dp / d, root, ignore_patterns, _cache=ignore_cache)
                 ]
             for fname in filenames:
                 if fname in _SKIP_FILES:
diff --git a/tests/test_detect.py b/tests/test_detect.py
index 118164dd..0c8ef484 100644
--- a/tests/test_detect.py
+++ b/tests/test_detect.py
@@ -542,6 +542,51 @@ def test_negation_ancestor_itself_reincluded(tmp_path):
     assert not _is_ignored(f, tmp_path, patterns)
 
 
+def test_negation_does_not_disable_directory_pruning(tmp_path, monkeypatch):
+    """A single `!` re-include must not switch off pruning of *unrelated* ignored dirs.
+
+    Regression: a blanket ``has_negation`` flag used to disable directory-level pruning
+    for EVERY ignored dir whenever any ``!`` pattern existed, so a single ``!docs/**``
+    made os.walk descend bin/, obj/, wwwroot/, generated/, … — a pathological slowdown
+    on large repos. Output stayed correct (the per-file ``_is_ignored`` filter still
+    excluded those files), so this guards the *walk* itself: the ignored dir must never
+    be descended, while the negation must still re-include its target.
+    """
+    import os
+    import graphify.detect as det
+
+    (tmp_path / ".graphifyignore").write_text("myignored/\n*.md\n!docs/**\n")
+    deep = tmp_path / "myignored" / "deep" / "deeper"
+    deep.mkdir(parents=True)
+    (deep / "junk.py").write_text("x = 1")
+    (tmp_path / "docs").mkdir()
+    (tmp_path / "docs" / "guide.md").write_text("# guide")
+    (tmp_path / "src").mkdir()
+    (tmp_path / "src" / "app.py").write_text("y = 2")
+
+    visited: list[str] = []
+    real_walk = os.walk
+
+    def tracking_walk(top, *args, **kwargs):
+        for dirpath, dirnames, filenames in real_walk(top, *args, **kwargs):
+            visited.append(dirpath)
+            yield dirpath, dirnames, filenames
+
+    monkeypatch.setattr(det.os, "walk", tracking_walk)
+    result = det.detect(tmp_path)
+
+    # The ignored (non-noise) dir must never be descended, despite the !docs/** negation.
+    assert not any("myignored" in Path(v).parts for v in visited), (
+        "ignored 'myignored/' was walked despite being ignored — the has_negation bypass regressed"
+    )
+    # Detection itself is unaffected: negation still re-includes docs/*.md, real source is
+    # found, and nothing leaks out of the ignored dir.
+    all_files = [p for cat in result["files"].values() for p in cat]
+    assert any(p.endswith("app.py") for p in all_files)
+    assert any(p.endswith("guide.md") for p in all_files)
+    assert not any("junk.py" in p for p in all_files)
+
+
 # Regression tests for #1087 - anchored patterns must not match basename deep in tree
 
 def test_anchored_dir_not_matched_at_depth(tmp_path):