stage2_worker: v2.1 — terminal failure states + sudo path fix

Three classes of silent failure converted to clean terminal states:

- Mistral timeout: previously left rows in zombie state (started_at set,
  failed_at null, attempts incremented past retry threshold, row invisible
  to selection query). Now sets failed_at with reason
  'mistral_timeout_after_300s'. Surfaced 2026-05-01 when 17 documents
  accumulated in this state during the Stage 3 saga deadlock incident.

- Mistral parse failure: run_mistral returns {'error': 'parse_failed'} on
  JSON decode failure but process_one wasn't checking, so empty orientation
  ('Active frames: . Frame relationships: ...') was shipped to Stage 3.
  This is F22 from the 2026-04-30 code review. Now sets failed_at with
  reason 'mistral_parse_failure'.

- Wedge recovery hammering: consecutive_failures was only reset on
  successful Ollama restart. With the sudo path bug (also fixed here),
  recovery always failed, so every subsequent failure re-attempted restart.
  Now resets the counter regardless and logs the failure visibly.

Also: subprocess.run now uses absolute paths (/usr/bin/sudo,
/bin/systemctl) instead of relying on PATH, fixing the 'No such file or
directory: sudo' error that broke Stage 2's recover_wedge() since
deployment. F45-adjacent — sudoers entries were added 2026-05-01 but the
PATH issue was masking that fix.

Worker version bumped to 2.1 to match Stage 3's resilience patch level.

scripts/stage2_worker.py

@@ -33,7 +33,7 @@ CHAR_LENGTH_THRESHOLD = 2000
 REQUEST_TIMEOUT = 300
 RETRY_ATTEMPTS = 2
 POLL_INTERVAL = 5
-WORKER_VERSION = "2.0"
+WORKER_VERSION = "2.1"
 
 TAXFREE_PROMPT = (
     "You are a metadata extraction system. Given a document, describe its content "
@@ -67,7 +67,10 @@ def write_heartbeat():
 
 def recover_wedge():
     log.warning("Mistral wedge detected — restarting Ollama")
-    subprocess.run(["sudo", "systemctl", "restart", "ollama"], capture_output=True)
+    result = subprocess.run(["/usr/bin/sudo", "/bin/systemctl", "restart", "ollama"], capture_output=True, text=True)
+    if result.returncode != 0:
+        log.error(f"Ollama restart failed (rc={result.returncode}): stdout={result.stdout!r} stderr={result.stderr!r}")
+        return False
     time.sleep(30)
     for _ in range(3):
         try:
@@ -146,6 +149,11 @@ def process_one(row):
         meta = run_mistral(full_text)
     except requests.exceptions.Timeout:
         log.warning(f"  Mistral timeout on {source}")
+        cur.execute(
+            "UPDATE stage_2_queue SET failed_at = NOW(), failure_reason = %s WHERE id = %s",
+            (f"mistral_timeout_after_{REQUEST_TIMEOUT}s", row_id)
+        )
+        pg.commit()
         pg.close()
         return False
     except Exception as e:
@@ -156,6 +164,16 @@ def process_one(row):
         pg.close()
         return False
 
+    if meta.get("error") == "parse_failed":
+        log.warning(f"  Mistral parse failure on {source}: {meta.get('raw', '')[:100]}")
+        cur.execute(
+            "UPDATE stage_2_queue SET failed_at = NOW(), failure_reason = %s WHERE id = %s",
+            ("mistral_parse_failure", row_id)
+        )
+        pg.commit()
+        pg.close()
+        return False
+
     frames = meta.get("active_frames", [])
     log.info(f"  Frames: {frames}")
 
@@ -209,8 +227,9 @@ def run():
                 if consecutive_failures >= 2:
                     log.warning("Multiple consecutive failures — checking for Mistral wedge")
                     recovered = recover_wedge()
-                    if recovered:
+                    if not recovered:
-                        consecutive_failures = 0
+                        log.error("Wedge recovery failed — continuing anyway")
+                    consecutive_failures = 0
                 time.sleep(10)
             else:
                 consecutive_failures = 0