Cite the cat-mode rule by quote, not line number, and test the quote exists

engine/hooks/unverified-tag-ledger/README.md

@@ -5,7 +5,7 @@ A well-formed `{{CAT-UNVERIFIED: <claim> -- cannot verify: <reason>}}` tag is a
 
 ## The gap it closes
 
-`cat-mode/SKILL.md:269` says:
+`cat-mode/SKILL.md` says:
 
 > Any hedge auto-runs prove-it in the same turn — a hedge is a trigger to
 > verify, never a place to stop.

engine/hooks/unverified-tag-ledger/claude_prompt_reminder.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 """Claude Code UserPromptSubmit hook: surface CAT-UNVERIFIED claims that
-earlier turns deferred and never settled. This is where cat-mode/SKILL.md:269
-gets teeth -- the Stop hook cannot block the turn that emits a tag without
+earlier turns deferred and never settled. This is where cat-mode/SKILL.md ("a hedge is a trigger to
+verify, never a place to stop") gets teeth -- the Stop hook cannot block the turn that emits a tag without
 deadlocking, so the reminder lands on the next prompt instead.
 
 How much it says is CATSTACK_UNVERIFIED_TAG_BEHAVIOR: off, stale (default),

engine/hooks/unverified-tag-ledger/claude_stop_check.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 """Claude Code Stop hook: record well-formed CAT-UNVERIFIED tags against the
 session, and refuse a turn that tags a claim without having run any
-verification tool (cat-mode/SKILL.md:269 -- a hedge is a trigger to verify).
+verification tool (cat-mode/SKILL.md -- a hedge is a trigger to verify).
 `stop_hook_active` releases that refusal so the rewrite turn can finish. With
 CATSTACK_UNVERIFIED_TAG_BEHAVIOR=do_not_emit it also refuses any reply that
 carries a tag. Fails open on read or parse errors.

engine/hooks/unverified-tag-ledger/detect.py

@@ -3,7 +3,7 @@
 
 Every other evidence hook treats `markers.well_formed_tags(message)` as
 equivalent to evidence and goes silent (see prove-it-ship-gate/detect.py).
-cat-mode/SKILL.md:269 says the opposite: "Any hedge auto-runs prove-it in the
+cat-mode/SKILL.md says the opposite: "Any hedge auto-runs prove-it in the
 same turn -- a hedge is a trigger to verify, never a place to stop." Nothing
 reconciled the two, so a correctly-formed tag was a free, unlogged exit.
 
@@ -229,7 +229,7 @@ def reminder(session_id: str, mode: str = DEFAULT_BEHAVIOR_MODE) -> str:
     lines = [
         "unverified-tag-ledger: "
         f"{len(open_rows)} CAT-UNVERIFIED claim(s) from earlier turns are still unverified.",
-        "cat-mode/SKILL.md:269 -- a hedge is a trigger to verify, never a place to stop. "
+        "cat-mode/SKILL.md -- a hedge is a trigger to verify, never a place to stop. "
         "The tag deferred these; it did not settle them.",
     ]
     for row in open_rows[:MAX_LISTED]:
@@ -275,7 +275,7 @@ def handback_offers(message: str) -> list[str]:
 def evaluate(payload: dict, mode: str | None = None) -> dict:
     """Record the turn, then decide whether it may end.
 
-    A tag earns its place only after an attempt. cat-mode/SKILL.md:269 asks for
+    A tag earns its place only after an attempt. cat-mode/SKILL.md asks for
     a verify in the SAME turn, so a tag emitted by a turn that ran no
     verification tool is a claim nobody tried to check, and that turn is
     refused. Requiring an attempt is not requiring success: run the check, and
@@ -350,7 +350,7 @@ def evaluate(payload: dict, mode: str | None = None) -> dict:
         return {"note": "", "block": (
             "unverified-tag-ledger: this turn tags a claim as unverified but ran no "
             f"verification tool. Untried claim(s): {claims}. "
-            "cat-mode/SKILL.md:269 -- a hedge is a trigger to verify, never a place to stop. "
+            "cat-mode/SKILL.md -- a hedge is a trigger to verify, never a place to stop. "
             "Run the check now (Bash/Read/Grep/Glob) and paste its output. The tag is for a "
             "check that was attempted and could not settle the claim, not for one nobody ran.")}
 
@@ -360,7 +360,7 @@ def evaluate(payload: dict, mode: str | None = None) -> dict:
         notes.append(
             f"unverified-tag-ledger: logged {len(fresh)} CAT-UNVERIFIED claim(s) against this "
             "session. They are deferred, not discharged, and will be raised again next turn "
-            "(cat-mode/SKILL.md:269).")
+            "(cat-mode/SKILL.md: a hedge is a trigger to verify).")
     return {"note": "\n".join(notes), "block": ""}
 
 

engine/hooks/unverified-tag-ledger/tests/test_hooks.py

@@ -17,6 +17,7 @@
 import io
 import json
 import os
+import re
 import sys
 import tempfile
 import unittest
@@ -33,6 +34,9 @@
     os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(HOOK))), ".env"),
 ])
 
+REPO_ROOT = os.path.dirname(os.path.dirname(os.path.dirname(HOOK)))
+CAT_MODE_PHRASE = "a hedge is a trigger to verify, never a place to stop"
+
 REAL_TAG_1 = (
     "{{CAT-UNVERIFIED: that it widened scope past the one session I gave it "
     "-- cannot verify: it never answered when asked twice; its "
@@ -167,9 +171,27 @@ def test_tag_after_a_real_attempt_is_logged_and_allowed(self) -> None:
     def test_tag_with_no_attempt_is_blocked(self) -> None:
         verdict = self.detect.evaluate(self.payload(REAL_TAG_1, tools=False))
         self.assertIn("ran no verification tool", verdict["block"])
-        self.assertIn("cat-mode/SKILL.md:269", verdict["block"])
+        self.assertIn("a hedge is a trigger to verify", verdict["block"])
         self.assertIn("widened scope", verdict["block"])
 
+    def test_the_quoted_rule_exists_in_the_cat_mode_skill_this_repo_owns(self) -> None:
+        skill = os.path.join(REPO_ROOT, "corpus", "skills", "cat-mode", "SKILL.md")
+        with open(skill, encoding="utf-8") as handle:
+            text = " ".join(handle.read().split())
+        self.assertIn(CAT_MODE_PHRASE, text)
+
+    def test_no_hook_source_cites_a_cat_mode_line_number(self) -> None:
+        stale = []
+        for name in sorted(os.listdir(HOOK)):
+            path = os.path.join(HOOK, name)
+            if not name.endswith((".py", ".md")) or not os.path.isfile(path):
+                continue
+            with open(path, encoding="utf-8") as handle:
+                for number, line in enumerate(handle, start=1):
+                    if re.search(r"SKILL\.md:\d+", line):
+                        stale.append(f"{name}:{number}")
+        self.assertEqual(stale, [])
+
     def test_blocked_turn_is_still_recorded(self) -> None:
         self.detect.evaluate(self.payload(REAL_TAG_1, tools=False))
         self.assertEqual(len(self.detect.outstanding(self.detect.read_ledger("s1"))), 1)
@@ -271,7 +293,7 @@ def test_reminder_names_the_claim_and_cites_the_rule(self) -> None:
         self.detect.record_turn("s1", REAL_TAG_1, set())
         text = self.detect.reminder("s1", "all")
         self.assertIn("widened scope", text)
-        self.assertIn("cat-mode/SKILL.md:269", text)
+        self.assertIn("a hedge is a trigger to verify", text)
         self.assertIn("never a place to stop", text)
 
     def test_reminder_is_silent_on_off(self) -> None: