fix: CodeRabbit follow-ups from #144 + #141 (incl. critical /api/v1 double-prefix)

Gradata/src/gradata/cli.py

@@ -1029,12 +1029,13 @@ def cmd_skill_export(args):
     Produces ``<output-dir>/<slug>/SKILL.md`` ready to drop into
     ``.claude/skills/`` or any Skills-aware harness.
     """
+    from gradata import Brain
     from gradata.enhancements.skill_export import export_skill, write_skill
 
     brain_root = _resolve_brain_root(args)
     lessons_path: Path | None = None
     try:
-        brain = _get_brain(args)
+        brain = Brain(brain_root)
         lessons_path = brain._find_lessons_path()
     except Exception:
         lessons_path = None

Gradata/src/gradata/cloud/sync.py

@@ -223,7 +223,7 @@ def sync_metrics(self, payload: TelemetryPayload) -> bool:
         # Backend mounts the metrics router under /api/v1 (see
         # cloud/app/main.py → app.include_router(router, prefix="/api/v1")
         # and cloud/app/routes/metrics.py → @router.post("/telemetry/metrics")).
-        result = self._post("/api/v1/telemetry/metrics", asdict(payload))
+        result = self._post("/telemetry/metrics", asdict(payload))
         if result is not None:
             self.config.last_sync_at = payload.sent_at
             save_config(self.brain_dir, self.config)
@@ -241,7 +241,7 @@ def contribute_corpus(self, anonymized_patterns: list[dict]) -> bool:
             return False
         # Backend mounts the corpus router under /api/v1 (same prefix as
         # telemetry — see cloud/app/main.py).
-        result = self._post("/api/v1/corpus/contribute", {"patterns": anonymized_patterns})
+        result = self._post("/corpus/contribute", {"patterns": anonymized_patterns})
         return result is not None
 
 

Gradata/src/gradata/enhancements/skill_export.py

@@ -99,7 +99,9 @@ def _format_skill_md(
     lines.append("---")
     lines.append(f"name: {name}")
     # Quote the description so colons / hashes inside don't break YAML.
-    safe_desc = description.replace('"', '\\"')
+    # Escape backslashes first, then quotes, so YAML stays valid even for
+    # Windows paths or descriptions containing literal `\\` sequences.
+    safe_desc = description.replace("\\", "\\\\").replace('"', '\\"')
     lines.append(f'description: "{safe_desc}"')
     lines.append("---")
     lines.append("")
@@ -196,7 +198,10 @@ def export_skill(
     rules = _parse_rules(Path(brain_root), lessons_path=lessons_path)
     rules = _filter_rules(rules, category)
     metas = _load_meta_principles(Path(brain_root)) if include_meta else []
-    desc = description.strip() if description else _auto_description(rules, slug)
+    if description and description.strip():
+        desc = description.strip()
+    else:
+        desc = _auto_description(rules, slug)
     if len(desc) > _DESC_MAX_LEN:
         desc = desc[: _DESC_MAX_LEN - 3] + "..."
     return _format_skill_md(slug, desc, rules, metas)

Gradata/src/gradata/hooks/implicit_feedback.py

@@ -162,6 +162,14 @@ def main(data: dict) -> dict | None:
         signal_types = {s["type"] for s in signals}
         has_negative = bool(signal_types & _NEGATIVE_SIGNAL_TYPES)
         has_approval = "approval" in signal_types
+        # Conflict resolution: when negative and approval coexist on the same
+        # message, negative wins. The user is critiquing — the polite "thanks"
+        # doesn't override "but this is wrong". Drop approval from emitted
+        # signals and suppress OUTPUT_ACCEPTED for this message.
+        if has_negative and has_approval:
+            signals = [s for s in signals if s["type"] != "approval"]
+            signal_types = {s["type"] for s in signals}
+            has_approval = False
         # Tacit acceptance: substantive follow-up with no negative signals. The
         # brain.correct() pipeline logs ~20x more CORRECTION than OUTPUT_ACCEPTED
         # because users rarely type "looks good" — silence is approval.

Gradata/src/gradata/hooks/inject_brain_rules.py

@@ -182,7 +182,14 @@ def _read_brain_prompt(brain_dir: Path) -> str | None:
         count=1,
     )
     # Limit to first GRADATA_WISDOM_MAX_RULES non-negotiable rules.
-    wisdom_max_rules = int(os.environ.get("GRADATA_WISDOM_MAX_RULES", "9"))
+    _raw_max = os.environ.get("GRADATA_WISDOM_MAX_RULES", "9")
+    try:
+        wisdom_max_rules = int(_raw_max)
+    except (ValueError, TypeError):
+        _log.warning(
+            "GRADATA_WISDOM_MAX_RULES=%r not an int — defaulting to 9", _raw_max
+        )
+        wisdom_max_rules = 9
     if wisdom_max_rules > 0:
         rule_lines = [ln for ln in text.split("\n") if ln.startswith("- ")]
         if len(rule_lines) > wisdom_max_rules:

Gradata/tests/test_cloud_sync.py

@@ -129,7 +129,7 @@ def test_sync_metrics_posts_when_enabled(self, tmp_path: Path):
         assert result is True
         mock_post.assert_called_once()
         call_path = mock_post.call_args[0][0]
-        assert call_path == "/api/v1/telemetry/metrics"
+        assert call_path == "/telemetry/metrics"
 
     def test_sync_metrics_updates_last_sync_at_on_success(self, tmp_path: Path):
         cfg = CloudConfig(sync_enabled=True, token="abc")
@@ -168,7 +168,7 @@ def test_contribute_corpus_posts_when_both_flags_set(self, tmp_path: Path):
 
         assert result is True
         mock_post.assert_called_once()
-        assert mock_post.call_args[0][0] == "/api/v1/corpus/contribute"
+        assert mock_post.call_args[0][0] == "/corpus/contribute"
 
 
 class TestConvenienceSync:

Gradata/tests/test_hooks_intelligence.py

@@ -465,7 +465,7 @@ def test_implicit_feedback_detects_challenge(tmp_path, monkeypatch):
     monkeypatch.setenv("GRADATA_BRAIN_DIR", str(tmp_path))
     with patch("gradata.hooks.implicit_feedback.emit_hook_event") as mock_emit:
         result = feedback_main({"message": "Are you sure that's correct? It doesn't look right."})
-    assert result is not None and "chal" in result["result"]
+    assert result == {"result": "[fb:chal]"}
     event_types = [call.args[0] for call in mock_emit.call_args_list]
     assert "IMPLICIT_FEEDBACK" in event_types
     signals = mock_emit.call_args_list[0].args[2]["signals"]
@@ -483,7 +483,7 @@ def test_implicit_feedback_emits_event(tmp_path):
         patch("gradata.hooks.implicit_feedback.emit_hook_event") as mock_emit,
     ):
         result = feedback_main({"message": "I told you not to do that, are you sure?"})
-    assert result is not None and result["result"].startswith("[fb:")
+    assert result == {"result": "[fb:rem,chal]"}
     event_types = [call.args[0] for call in mock_emit.call_args_list]
     assert "IMPLICIT_FEEDBACK" in event_types
 

Gradata/tests/test_skill_export.py

@@ -123,6 +123,34 @@ def test_double_quotes_in_description_are_escaped(self, tmp_path: Path) -> None:
         # Ensure the quote is backslash-escaped so YAML stays valid
         assert r'description: "He said \"hi\" loudly"' in text
 
+    def test_backslash_in_description_is_escaped(self, tmp_path: Path) -> None:
+        # Windows path with literal backslashes must produce an escaped YAML string.
+        text = export_skill(tmp_path, name="demo", description=r"C:\Users\foo")
+        assert r'description: "C:\\Users\\foo"' in text
+
+    def test_backslash_and_quote_combined(self, tmp_path: Path) -> None:
+        # Backslashes must be escaped BEFORE quotes so we don't accidentally
+        # produce \\" sequences that break YAML.
+        text = export_skill(tmp_path, name="demo", description=r'path C:\a "b"')
+        assert r'description: "path C:\\a \"b\""' in text
+
+    def test_multiline_literal_newline_in_description(self, tmp_path: Path) -> None:
+        # A literal `\n` (two chars: backslash + n) must round-trip as `\\n`.
+        text = export_skill(tmp_path, name="demo", description=r"line1\nline2")
+        assert r'description: "line1\\nline2"' in text
+
+    def test_whitespace_only_description_falls_back_to_auto(
+        self, tmp_path: Path
+    ) -> None:
+        _write_lessons(tmp_path, SAMPLE_LESSONS)
+        text = export_skill(tmp_path, name="demo", description="   \t\n  ")
+        # Auto description must not be empty and must not be just whitespace.
+        import re
+
+        m = re.search(r'^description: "(.*)"$', text, flags=re.MULTILINE)
+        assert m is not None, "description line missing"
+        assert m.group(1).strip() != "", "whitespace-only desc should fall back to auto"
+
     def test_category_filter_narrows_output(self, tmp_path: Path) -> None:
         _write_lessons(tmp_path, SAMPLE_LESSONS)
         text = export_skill(tmp_path, name="demo", category="DRAFTING")