Update EVAL_RESULTS-frontier.md: difficulty wording and caveats

EVAL_RESULTS-frontier.md

@@ -20,11 +20,10 @@ Logs are retained. Model used objdump/qemu/WSL via Bash tool and verified by run
 ## 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 no-tools-requested 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 our published claims: single-shot no-tools-requested frontier is NOT reliable on binary ID (58% here; our separate local baselines are different samples — we keep the numbers separate).
 - CONFIRMS the flag-hunt envs are agent-SOLVABLE with tools in ~1 min → warm-up/oracle checks, not strong RL signal as shipped.
 - Our edge is NOT "frontier can't solve it" — it is **verified deterministic oracles at scale (10,251 items / 29 envs /
-  6 ISAs+PE), zero-leak construction, and per-item provenance** — exactly what DataVendor/HUD quality gates check.
+  6 ISAs+PE), automated exact-token leak scan 0/10,251, and per-item provenance** — exactly what DataVendor/HUD quality gates check.
 - Buyer-facing use: we can quote frontier reference scores unprompted (quality-gate pre-answer). A buyer's own model can be run on a 10-env probe as a quoted calibration step before purchase.
 
 ## C) Hard-trial agentic spot check — 20 sampled hard items (2026-10-30)
@@ -67,6 +66,6 @@ independent Bernoulli trials and are wide at this sample size):
 Honest reading: these 10 sampled families are **practice level for a frontier agent**: each family's Wilson lower bound
 (51%) is above the planned 2–30% RL band, and the task-level `20/20` gives a pooled lower bound of about 84%. The other
 9 families (`bitmath`, `deflate_bytes`, `fnv_hash`, `glibc_ldtrace`, `ldl_trace_x86`, `posix_collate`, `rpath_runpath`,
-`signal_rc`, `tz_fold_gap`) have not been run agentically. The local 3-bit-quantized open-weight `14/99` number (text-only prompt with no binary/file access, no tools, single-turn, temperature 0.7, max_tokens=600) remains headroom evidence, not an agentic difficulty claim. To sell hard tasks as RL-band material, families need to be made
+`signal_rc`, `tz_fold_gap`) have not been run agentically. The local 3-bit-quantized open-weight `14/99` number (text-only prompt with no binary/file access, no tools, single-turn, temperature 0.7, max_tokens=600, seeds 1000–1007, whole-reply normalized matching) remains headroom evidence, not an agentic difficulty claim. To sell hard tasks as RL-band material, families need to be made
 harder for agentic use: remove the direct "run the exact command" shortcut, add multi-step or cross-binary goals,
 strip/obfuscate more, and require answers that are not printed by a single reference command.