From 4e320033ad74c3c88fd406996dd99efa406d1ced Mon Sep 17 00:00:00 2001
From: ciregenz <ciregenz@berkeley.edu>
Date: Sat, 6 Jun 2026 04:49:19 -0700
Subject: [PATCH] [eric] browser: prune abandoned-navigate detours at distill
 so recorded skills and route hints stay clean

---
 backend/apps/agents/browser/browser_skills.py | 18 ++++++++++-
 backend/tests/test_browser_skills.py          | 32 +++++++++++++++++++
 2 files changed, 49 insertions(+), 1 deletion(-)

diff --git a/backend/apps/agents/browser/browser_skills.py b/backend/apps/agents/browser/browser_skills.py
index f024ac50..515f45e0 100644
--- a/backend/apps/agents/browser/browser_skills.py
+++ b/backend/apps/agents/browser/browser_skills.py
@@ -309,7 +309,23 @@ def distill_steps(action_log: list[dict]) -> list[dict]:
             productive_count += 1
     if productive_count == 0:
         return []
-    return steps
+    return _prune_detours(steps)
+
+
+def _prune_detours(steps: list[dict]) -> list[dict]:
+    """Drop an abandoned-page detour: a BrowserNavigate whose page was never
+    acted on because the very next step navigates somewhere else. Conservative
+    on purpose, only consecutive navigates qualify (if a page had been used,
+    a Type/Click/etc would sit between them), so a needed step is never removed.
+    This keeps a wrong-turn (e.g. the wrong profile, then the right one) out of
+    a recorded macro without any reachability guesswork."""
+    out: list[dict] = []
+    for i, s in enumerate(steps):
+        nxt = steps[i + 1] if i + 1 < len(steps) else None
+        if s.get("tool") == "BrowserNavigate" and nxt is not None and nxt.get("tool") == "BrowserNavigate":
+            continue  # this navigate's page was abandoned immediately; skip it
+        out.append(s)
+    return out
 
 
 def first_unsafe_step(steps: list[dict]) -> tuple[int, str]:
diff --git a/backend/tests/test_browser_skills.py b/backend/tests/test_browser_skills.py
index 8cc532a2..95a2eecf 100644
--- a/backend/tests/test_browser_skills.py
+++ b/backend/tests/test_browser_skills.py
@@ -665,3 +665,35 @@ def test_route_hint_adoption_matching(_isolated_skills):
     # the Message and Send clicks did not run
     assert adopted[0] and adopted[1] and adopted[3]
     assert not adopted[2] and not adopted[4]
+
+
+# --- conservative detour pruning ---------------------------------------------
+def test_distill_prunes_abandoned_navigate_detour():
+    # wrong profile opened (navigate), abandoned for a search (navigate), then
+    # the right profile + the real productive steps. The first navigate is a
+    # detour: nothing acted on its page before the next navigate.
+    log = [
+        {"tool": "BrowserNavigate", "input": {"url": "https://x.com/in/wrong"}, "ok": True},
+        {"tool": "BrowserNavigate", "input": {"url": "https://x.com/search?q=tyler"}, "ok": True},
+        {"tool": "BrowserClickIndex", "input": {"index": 1}, "ok": True,
+         "clicked_role": "link", "clicked_name": "Tyler Chen"},
+        {"tool": "BrowserType", "input": {"selector": "#msg", "text": "hi"}, "ok": True},
+    ]
+    steps = sk.distill_steps(log)
+    urls = [s["params"].get("url") for s in steps if s["tool"] == "BrowserNavigate"]
+    assert urls == ["https://x.com/search?q=tyler"]  # the abandoned wrong nav dropped
+    assert [s["tool"] for s in steps] == ["BrowserNavigate", "BrowserClickByName", "BrowserType"]
+
+
+def test_distill_keeps_navigate_that_was_acted_on():
+    # a navigate followed by a real action on that page is NOT a detour.
+    log = [
+        {"tool": "BrowserNavigate", "input": {"url": "https://x.com/form"}, "ok": True},
+        {"tool": "BrowserType", "input": {"selector": "#q", "text": "hi"}, "ok": True},
+        {"tool": "BrowserNavigate", "input": {"url": "https://x.com/results"}, "ok": True},
+        {"tool": "BrowserClickIndex", "input": {"index": 2}, "ok": True,
+         "clicked_role": "button", "clicked_name": "Go"},
+    ]
+    steps = sk.distill_steps(log)
+    urls = [s["params"].get("url") for s in steps if s["tool"] == "BrowserNavigate"]
+    assert urls == ["https://x.com/form", "https://x.com/results"]  # both kept