fix: vision OCR receipt extraction — skip second LLM call, fix total truncation

receipt_parser: change _ocr_image_vision() to extract structured JSON
{vendor,amount,date,time,category} directly from the image instead of
transcribing raw text, so the downstream LLM extraction step is
unnecessary and the two-step error-compounding is eliminated.

expenses_agent: add _match_category() helper to map vision category
labels to expense product names via substring/fuzzy match; add fast
path in _parse_receipt_text() that detects pre-extracted vision JSON
(text starts with '{') and skips the second LLM submit call entirely.
Fix text[:2000] truncation that discarded receipt totals — now keeps
first 1500 + last 1500 chars of long receipts so the grand total at
the bottom is always included.

tests: fix stale test_act_enters_awaiting_confirmation_on_first_pass
(confirmation gate was removed); add TestMatchCategory and three new
tests for the vision JSON fast path and LLM fallthrough.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

agent_service/tools/receipt_parser.py

@@ -98,7 +98,13 @@ def _ocr_image(data: bytes, filename: str) -> str:
 
 
 def _ocr_image_vision(data: bytes, filename: str, ollama_url: str, model: str) -> str:
-    """Use an Ollama vision model to read a receipt image."""
+    """Use an Ollama vision model to extract receipt data directly as JSON.
+
+    Returns a JSON string {vendor, amount, date, time, category} so the
+    expenses agent can skip the second LLM extraction step entirely.
+    Returns empty string on any failure so the caller falls back to Tesseract.
+    """
+    import json as _json
     try:
         import ollama as _ollama
         client = _ollama.Client(host=ollama_url)
@@ -107,22 +113,41 @@ def _ocr_image_vision(data: bytes, filename: str, ollama_url: str, model: str) -
             messages=[{
                 'role': 'user',
                 'content': (
-                    'This is a photo of a paper receipt. '
-                    'Transcribe ALL text exactly as it appears on the receipt. '
-                    'Preserve every line in order: store name, address, date, time, '
-                    'each line item with price, subtotal, tax, tip if present, and '
-                    'the final total. Output the raw text only — no commentary, '
-                    'no markdown, no explanations.'
+                    'This is a photo of a receipt. Extract these fields:\n'
+                    '- vendor: the store or restaurant name\n'
+                    '- amount: the FINAL total the customer paid. Look for a line '
+                    'labeled "Total", "Grand Total", "Amount Due", or "Balance Due". '
+                    'Do NOT use subtotal, tax, or tip. Return 0 if you cannot find '
+                    'a clear final total.\n'
+                    '- date: transaction date in YYYY-MM-DD format\n'
+                    '- time: transaction time in HH:MM 24-hour format, or null\n'
+                    '- category: one word describing the expense type — one of: '
+                    'meals, fuel, hotel, office, transport, other\n\n'
+                    'Return ONLY a valid JSON object, no commentary, no markdown:\n'
+                    '{"vendor":"...","amount":0.00,"date":"YYYY-MM-DD",'
+                    '"time":"HH:MM or null","category":"..."}'
                 ),
                 'images': [data],
             }],
         )
         if isinstance(response, dict):
-            text = (response.get('message', {}).get('content') or '').strip()
+            raw = (response.get('message', {}).get('content') or '').strip()
         else:
-            text = (response.message.content or '').strip()
-        logger.debug('Vision OCR %s (%s): %d chars', filename, model, len(text))
-        return text
+            raw = (response.message.content or '').strip()
+
+        # Must contain a JSON object, not prose
+        first, last = raw.find('{'), raw.rfind('}')
+        if first == -1 or last <= first:
+            logger.warning('Vision OCR %s: model returned prose, falling back to Tesseract',
+                           filename)
+            return ''
+        json_str = raw[first:last + 1]
+        parsed = _json.loads(json_str)
+        if 'amount' not in parsed:
+            logger.warning('Vision OCR %s: JSON missing amount field, falling back', filename)
+            return ''
+        logger.debug('Vision OCR %s (%s): extracted JSON ok', filename, model)
+        return json_str
     except ImportError:
         logger.warning('ollama package not installed — vision OCR unavailable for %s', filename)
         return ''