feat(learning-harness): fix dates and the post-cutoff slice

The catalog generator fills `fix_date` from the recipe at the precision the
recipe has -- a full ISO day when one was recorded, otherwise the year.
Every harvested recipe carries a year and almost none carries a day, and
writing `-01-01` would turn a year into a claim about a month. A row counts
as post-cutoff only when its earliest possible date is after the cutoff, so
a coarse date can under-claim the clean slice but never over-claim it.

Each family now reports the same numbers over the rows fixed after the
cutoff, beside the whole number and never instead of it, with the undated
count on its own line. Without a configured cutoff every scoreboard says so
rather than inventing one: guessing a model's training cutoff and then
publishing a contamination number from the guess would be worse than
reporting none.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01T7XD7x5KonMdmGRkp63EQG

benchmarks/pairs/catalog_gen.py

@@ -93,6 +93,9 @@ def catalog_text(slice_name: str, recipes: list[dict[str, object]]) -> str:
             lines.append(f'reviewer = "{_q(reviewer)}"')
         if recipe.get("known_limit"):
             lines.append(f'known_limit = "{_q(recipe["known_limit"])}"')
+        fix_date = _fix_date(recipe)
+        if fix_date:
+            lines.append(f'fix_date = "{_q(fix_date)}"')  # Req 10.5: the post-cutoff slice needs a date to sort by
         for relpath, context_vuln, context_fixed in context_rel(recipe):
             lines += [
                 "",
@@ -123,6 +126,19 @@ def catalog_text(slice_name: str, recipes: list[dict[str, object]]) -> str:
     return "\n".join(lines)
 
 
+def _fix_date(recipe: dict[str, object]) -> str:
+    """The date of the fix at the precision the recipe recorded it: a full ISO day, or the year, or nothing.
+
+    Every harvested recipe carries a year; almost none carries a day, and inventing `-01-01` would turn a year into
+    a claim about a month. The post-cutoff comparison reads the earliest possible date of whatever precision is
+    here, so a coarse date can only ever under-claim (learning-harness Req 10.5)."""
+    recorded = str(recipe.get("fix_date", "") or "")
+    if recorded:
+        return recorded
+    year = recipe.get("year")
+    return str(year) if isinstance(year, int) and 1990 <= year <= 2100 else ""
+
+
 _DEFAULT_TIER = {"vibe-py": "seeded", "vfc-js": "advisory", "agent-vfc": "title", "vfc": "advisory", "sast": "advisory"}