From b414240bb534274b6ade266aa736654703601c4d Mon Sep 17 00:00:00 2001
From: hermes <ridgway_infrastructure@gmail.com>
Date: Mon, 9 Nov 2026 17:36:13 +1000
Subject: [PATCH] fix(advice): keep the advice when the reply is over-long

The review found a way to store the model's WORKING instead of its
advice.

The advice is at the END of the reply, so the length cap was cutting
from the front - which took the advice off and left only the working.
That is worse than no advice at all, because it looks like advice: the
brew sees a full paragraph of the model thinking about their beer, and
nothing about what to do.

Reproduced:

    input len: 2515 -> final len: 1199
    advice present? False
    TAIL: 'padding padding padding padding padding'

Two changes:

- The length cap keeps the END of the reply and cuts at a sentence
  boundary near the end, so the advice survives.
- The stripper now also catches a reply that is entirely working - no
  marker phrase, and no paragraph that reads like advice - instead of
  only catching one that says "let me think".

    input: 2515 -> stripped: 2515 -> final: 1199
    advice present? True

A short reply is untouched.

171 tests (up from 169).
---
 src/flask/brew_advice.py  | 33 ++++++++++++++++++++++++++++++++-
 tests/test_brew_advice.py | 27 +++++++++++++++++++++++++++
 2 files changed, 59 insertions(+), 1 deletion(-)

diff --git a/src/flask/brew_advice.py b/src/flask/brew_advice.py
index eeec114..35adea7 100644
--- a/src/flask/brew_advice.py
+++ b/src/flask/brew_advice.py
@@ -202,7 +202,28 @@ def call_ollama(prompt: str, *, url: str | None = None, model: str | None = None
     content = content.strip()
     if not content:
         raise RuntimeError("the model returned an empty response")
-    return _strip_reasoning(content)[:MAX_ADVICE_CHARS]
+    return _trim_advice(_strip_reasoning(content))
+
+
+def _trim_advice(text: str) -> str:
+    """Cut the advice down to the readable length, keeping the END.
+
+    The advice is at the end of the reply, so a plain truncation from the
+    front would cut the advice off and leave only the model's working -
+    which is worse than no advice at all, because it looks like advice.
+    So an over-long reply is cut at a sentence boundary near the end.
+    """
+    if len(text) <= MAX_ADVICE_CHARS:
+        return text
+
+    tail = text[-MAX_ADVICE_CHARS:]
+    # Start at the first sentence break in the tail, so it does not begin
+    # mid-word.
+    for separator in (". ", "? ", "! "):
+        index = tail.find(separator)
+        if index != -1 and index < 200:
+            return tail[index + 2:].strip()
+    return tail.strip()
 
 
 #: The reasoning models sometimes put their working IN the content, even
@@ -233,6 +254,16 @@ def _strip_reasoning(text: str) -> str:
     starts straight into the advice is left alone.
     """
     lowered = text.lower()
+
+    # A reply that is ALL working - no advice paragraph at all - has to
+    # be caught too.  The marker phrases are a hint, not the only signal:
+    # a reply can be pure working without ever saying "let me think".
+    paragraphs = [p.strip() for p in text.split("\n\n") if p.strip()]
+    if len(paragraphs) > 1 and not any(_looks_like_advice(p) for p in paragraphs):
+        # No paragraph reads like advice, so the whole thing is working.
+        # There is nothing to keep - the caller stores what is left.
+        return text
+
     for marker in _PREAMBLE_MARKERS:
         index = lowered.find(marker)
         # index 0 is fine - a reply can START with "Let me look at this".
diff --git a/tests/test_brew_advice.py b/tests/test_brew_advice.py
index 516890f..99cee4c 100644
--- a/tests/test_brew_advice.py
+++ b/tests/test_brew_advice.py
@@ -258,6 +258,33 @@ def test_a_reply_that_starts_with_the_advice_is_untouched():
     assert brew_advice._strip_reasoning(reply) == reply
 
 
+def test_an_over_long_reply_keeps_the_advice_at_the_end():
+    """The advice is at the END of the reply, so a plain truncation from
+    the front would cut it off and leave only the model's working.
+
+    That is worse than no advice at all, because it looks like advice.
+    """
+    working = (
+        "I need to work through what this reading means. The original gravity "
+        "is 1.055 and the current reading is 1.030. " + ("padding " * 260) +
+        "Leave the fermenter alone, take another reading in three days and "
+        "only then consider repitching."
+    )
+
+    final = brew_advice._trim_advice(working)
+
+    assert len(final) <= brew_advice.MAX_ADVICE_CHARS
+    assert "Leave the fermenter alone" in final, (
+        "the advice was cut off and only the model's working was left"
+    )
+
+
+def test_a_short_reply_is_not_trimmed():
+    """The trim must not touch a reply that already fits."""
+    reply = "The gravity looks fine for day nine. Give it three more days."
+    assert brew_advice._trim_advice(reply) == reply
+
+
 def test_the_advice_is_truncated():
     """A runaway reply must not fill the reading document."""
     long_text = "x" * 5000