benchmark final wording

EVAL_RESULTS-frontier.md

@@ -19,6 +19,8 @@ our deterministic graders catch. Raw logs: C:\RE-Lab\tmp\fr_item_out_1..12.json,
 Logs: C:\RE-Lab\tmp\fr_ag_out_*.json; prompts fr_ag_*.txt. Model used objdump/qemu/WSL via Bash tool and verified by running.
 
 ## What this means (honest reading)
+UPDATE same day: full pass@5 on all 29 envs now published — see **EVAL_RESULTS-passk.md** (Opus 5.5: 29/29 envs,
+145/145 attempts). Section B above was the pilot of that harness; its "3/3" is subsumed by the full result.
 - CONFIRMS our published claims: single-shot frontier is NOT reliable on binary ID (58% here; our local-model 0/5 is
   a different, weaker regime — we keep both numbers separate).
 - CONFIRMS the flag-hunt envs are agent-SOLVABLE with tools in ~1 min → good RL signal: reward reachable but not free.