no-mistakes(review): Split retired-alias audit into fail vs warn; fix key claim

audit-hermes-config.py

@@ -232,24 +232,35 @@ def audit(path):
         f"custom_providers[0].base_url must end with /v1 (got {cp.get('base_url')!r})",
     )
 
-    # --- No raw or retired model names (Rule 7/8 spirit) ---
-    # Any retired/raw name found by the derivation above is a hard failure: such a config gets
-    # 400 Invalid model name (or lands in the wrong pool) at runtime. gpu-light/gemma-4-12b/
-    # crew-auto were retired 2026-09-12; raw model names must use stable aliases instead.
-    bad_model_names = {
-        "gemma-4-12b": "gpu-vision",
+    # --- Retired/raw model names (Rule 7/8 spirit) ---
+    # The audit's job is to catch configs that are BROKEN, not to enforce a style preference.
+    # NON-RESOLVING names (removed 2026-09-12, verified 400/403 via live LiteLLM) must hard-FAIL:
+    #   gpu-light -> gpu-vision ; gemma-4-12b -> gpu-vision
+    #   crew-auto -> syslog-auto (its 64K cap is retired; no cap in force) ; ornith-1.0-35b -> strix-moe
+    # RESOLVING names (verified 200) are discouraged but working, so they only WARN:
+    #   qwen3.6-27B-code -> gpu-dense ; qwen3.6-35B-udq4 -> strix-moe
+    # Failing a working alias would reject valid configs - the exact defect this change fixes.
+    non_resolving = {
         "gpu-light": "gpu-vision",
+        "gemma-4-12b": "gpu-vision",
         "crew-auto": "syslog-auto (its 64K cap is retired; no cap in force)",
-        "qwen3.6-27B-code": "gpu-dense",
-        "qwen3.6-35B-udq4": "strix-moe",
         "ornith-1.0-35b": "strix-moe",
     }
+    raw_but_live = {
+        "qwen3.6-27B-code": "gpu-dense",
+        "qwen3.6-35B-udq4": "strix-moe",
+    }
     for field_path, value in _iter_model_values(cfg):
-        if value in bad_model_names:
+        if value in non_resolving:
             check(
                 False,
                 "Rule 7/8",
-                f"{field_path} = {value!r} is retired/raw (2026-09-12 sweep) — use {bad_model_names[value]}",
+                f"{field_path} = {value!r} is retired and no longer resolves (2026-09-12) — use {non_resolving[value]}",
+            )
+        elif value in raw_but_live:
+            warn(
+                "Rule 7/8",
+                f"{field_path} = {value!r} is a raw-but-live model name — prefer the stable alias {raw_but_live[value]}",
             )
 
     # --- Report ---

litellm-api-keys.prose.md

@@ -67,7 +67,7 @@ description: >
 4. **If action == "create"**:
    - Generate new key with key_alias: "{agent_name}" (e.g., "tanko" — bare name, no date)
    - Set metadata: { "agent": "{agent_name}", "purpose": "agent-inference" }
-   - Duration is null (permanent) — inherited from litellm default_key_generate_params
+   - Duration is whatever the caller passes; NO default enforcement exists today (CT 116 `litellm_config.yaml` has no `default_key_generate_params` block, and a key with no explicit models returns an empty models list). Agent keys are permanent by policy, not by that block. OPEN policy question: should agent keys expire by default? (captain security-policy decision, raised separately.)
    - Set models: read the live key-scoped set rather than hardcoding one — `/v1/models` is key-scoped,
      and the authoritative registry is CT 116 `/opt/inference-harness/litellm_config.yaml`. Do not add
      retired names (`gemma-4-12b`, `gpu-light`, `crew-auto` — all retired 2026-09-12).

tests/test_audit_hermes_config_alias.py

@@ -132,17 +132,20 @@ def test_retired_alias_in_custom_providers_is_rejected(tmp_path):
     assert "RESULT: FAIL" in out
 
 
-def test_raw_alias_is_rejected(tmp_path):
-    """Raw-but-live model names must fail too, pointing at the stable alias."""
+def test_raw_but_live_alias_warns_but_passes(tmp_path):
+    """Raw-but-live names resolve (200), so they warn only; failing them rejects valid configs."""
     code, out = _run_config(
         tmp_path,
         "raw-qwen.yaml",
-        BASE.format(alias="gpu-vision").replace("default: syslog-auto", "default: qwen3.6-27B-code"),
+        BASE.format(alias="gpu-vision").replace(
+            "delegation:\n  provider: harness",
+            "delegation:\n  provider: harness\n  model: qwen3.6-27B-code",
+        ),
     )
-    assert code == 1, out
-    assert "model.default = 'qwen3.6-27B-code'" in out
-    assert "gpu-dense" in out
-    assert "RESULT: FAIL" in out
+    assert code == 0, out
+    assert "delegation.model = 'qwen3.6-27B-code' is a raw-but-live model name" in out
+    assert "prefer the stable alias gpu-dense" in out
+    assert "RESULT: PASS" in out
 
 
 def test_retired_alias_in_fallback_providers_is_rejected(tmp_path):