opencode-cache-engine 0.4.3 → 0.4.5

package/README.md

@@ -31,8 +31,8 @@ The plugin currently has four cache-policy families:
 * **GPT-5.6 and later** — documented cache-key/options metadata, with prompt text
   unchanged. GPT-6 and future 5.6+/6+/7+ versions resolve through the same
   documented boundary.
-* **GLM-5.3** — narrow, content-preserving `<env>` relocation and diagnostics.
-* **MiMo-V2.6** — narrow, content-preserving `<env>` relocation and diagnostics.
+* **GLM-5.3 and later** — GLM implicit-cache baseline and diagnostics; GLM-5.3 additionally uses a narrow, content-preserving `<env>` relocation overlay.
+* **MiMo V2.6 and later** — MiMo implicit-cache baseline and diagnostics; MiMo V2.6 Flash/Pro additionally use a narrow, content-preserving `<env>` relocation overlay.
 
 Family classification is not hard-coded in the runtime. A pure policy registry
 and resolver in `src/cache-policy-core.mjs` returns a structured result
@@ -42,7 +42,7 @@ the single runtime source of policy classification. The first-party research
 behind each registry entry is recorded in
 [docs/cache-policy-inventory.md](docs/cache-policy-inventory.md).
 
-For both MiMo-V2.6 and GLM-5.3, CacheEngine adds its deterministic
+For the MiMo V2.6-and-later family and the GLM-5.3-and-later family, CacheEngine adds its deterministic
 `x-session-id` request header only when OpenCode identifies the actual provider
 as `openrouter`. It does not add that OpenRouter-specific header for
 non-OpenRouter providers; direct provider endpoints retain their provider-native
@@ -205,6 +205,12 @@ GLM-5.3 and MiMo-V2.6 use the only prompt-text transformation in the current
 plugin: a narrow, content-preserving relocation of the identifiable `<env>`
 block for the eligible model family.
 
+Since v0.4.4 the GLM **family baseline** and the GLM-5.3 **overlay** are
+separate. A resolved GLM-5.3-and-later model inherits the implicit-cache baseline
+and the non-mutating GLM diagnostics/transport, but the `<env>` relocation below
+is a GLM-5.3-specific overlay and is **not** inherited by a newer GLM merely
+because its version number is higher.
+
 The plugin identifies OpenCode's volatile `<env>` section and moves it to the **tail of the system prompt**.
 
 Conceptually:
@@ -280,19 +286,33 @@ whether a header was already present or added, and provider-identity changes
 (`glm_provider_changed`). They do not record the header value.
 
 
-## MiMo-V2.6 (Flash / Pro)
+## MiMo V2.6 and later
 
 ### Policy: prefix stability + OpenRouter session affinity
 
-MiMo-V2.6 is Xiaomi's current model family. The plugin targets exactly two
-identifiers:
+MiMo-V2.6 is Xiaomi's current model family. Since v0.4.5 the plugin separates
+three concerns:
+
+* **Family baseline** — the implicit-cache baseline plus cached-token telemetry,
+  provider-change/prefix diagnostics, and OpenRouter session affinity. This
+  applies to the documented V2.6 Flash and Pro identifiers, the
+  `mimo-v2.6-pro-ultraspeed` mode id, and any future MiMo generation after V2.6.
+* **Validated overlay** — the `<env>` relocation described below, registered only
+  for MiMo V2.6 Flash/Pro.
+* **Transport affinity** — OpenRouter `x-session-id`, gated on the actual
+  `openrouter` provider identity.
+
+The documented V2.6 identifiers are:
 
 * `xiaomi/mimo-v2.6-flash` / `mimo-v2.6-flash`
 * `xiaomi/mimo-v2.6-pro` / `mimo-v2.6-pro`
+* `xiaomi/mimo-v2.6-pro-ultraspeed` / `mimo-v2.6-pro-ultraspeed`
 
 Detection also tolerates `provider/model` shapes where `api.id` contains those
-slugs. It deliberately does **not** match `mimo-v2.5`, `mimo-v2.5-pro`,
-`mimo-v2.6-pro-ultraspeed`, or unrelated MiMo models.
+slugs. Future generations after V2.6 (for example `mimo-v2.7-*`) resolve to the
+family baseline so a new model remains usable even when its exact id is unknown.
+`mimo-v2.5`, `mimo-v2.5-pro`, `mimo-v2`, and undocumented V2.6 variants such as
+`mimo-v2.6-flashx` remain neutral.
 
 ### Implicit context caching
 
@@ -307,9 +327,9 @@ Implicit caching is the default assumption.
 
 ### Environment-block stabilization
 
-MiMo uses the same narrow, content-preserving transformation as GLM-5.3: the
-identifiable volatile `<env>` block is relocated to the **tail** of the single
-system string. Contents are preserved byte-for-byte; only position changes. This
+MiMo V2.6 Flash/Pro use the same narrow, content-preserving transformation as
+GLM-5.3: the identifiable volatile `<env>` block is relocated to the **tail** of
+the single system string. Contents are preserved byte-for-byte; only position changes. This
 keeps the large reusable prefix stable when only the environment/date changes.
 
 The transformation is applied only when:
@@ -343,12 +363,13 @@ Explicit telemetry events:
 
 ### OpenRouter session affinity
 
-For MiMo-V2.6 requests whose actual OpenCode provider identity is `openrouter`,
-CacheEngine adds its existing deterministic, session-scoped `x-session-id`
-request header. If a case-insensitive `x-session-id` already exists in model or
-plugin headers, CacheEngine preserves it and does not replace it. Eligibility
-uses both the MiMo-V2.6 family and the actual provider identity; a matching model
-slug on another endpoint is not enough.
+For MiMo V2.6-and-later requests whose actual OpenCode provider identity is
+`openrouter`, CacheEngine adds its existing deterministic, session-scoped
+`x-session-id` request header. If a case-insensitive `x-session-id` already
+exists in model or plugin headers, CacheEngine preserves it and does not replace
+it. Eligibility uses both the MiMo V2.6-and-later family and the actual provider
+identity; a matching model slug on another endpoint is not enough. This is
+transport affinity, separate from the V2.6 Flash/Pro `<env>` overlay.
 
 For Xiaomi's direct endpoint and every other non-OpenRouter provider, CacheEngine
 does not add its OpenRouter-specific `x-session-id`. Direct provider endpoints
@@ -428,8 +449,8 @@ and are reported diagnostically; the message content is left untouched.
 | ------------- | --------- | ------------------- | ----------------------- | -------------------------- | -------------------- |
 | DeepSeek | `deepseek` (V4-and-later family + passive fallback) | No | No | None | provider `cache.read` / `cache.write` |
 | GPT-5.6 and later | version boundary `gpt-<major>[.<minor>] ≥ 5.6` on OpenAI-ish endpoints (includes GPT-6) | No | Yes: `prompt_cache_key` + options | None | provider cache tokens |
-| GLM-5.3 | `glm-5.3*` | Yes, narrowly (`<env>` tail) | No provider cache key | `x-session-id` on OpenRouter only | provider cache tokens (GLM ratio) |
-| MiMo-V2.6 | Flash / Pro only | Yes, narrowly (`<env>` tail) | No: implicit caching only | `x-session-id` on OpenRouter only | `cached_tokens / prompt_tokens` |
+| GLM-5.3 and later | `glm-5.3+` | Yes, narrowly (`<env>` tail) on GLM-5.3 only | No provider cache key | `x-session-id` on OpenRouter only | provider cache tokens (GLM ratio) |
+| MiMo V2.6 and later | `mimo-v2.6+` (family) | Yes, narrowly (`<env>` tail) on V2.6 Flash/Pro only | No: implicit caching only | `x-session-id` on OpenRouter only | `cached_tokens / prompt_tokens` |
 
 `x-session-id` is an HTTP affinity header, not a provider cache key or
 cache-control field. Non-OpenRouter endpoints do not receive CacheEngine's
@@ -941,14 +962,15 @@ neutral
 The model detector recognizes:
 
 * DeepSeek model/provider identifiers
-* GPT-5.6 variants
-* GLM-5.3 variants
-* MiMo-V2.6 Flash and Pro (`xiaomi/mimo-v2.6-flash`, `mimo-v2.6-pro`, ...)
+* GPT-5.6-and-later variants
+* GLM-5.3-and-later variants
+* MiMo V2.6-and-later family (`mimo-v2.6-flash`, `mimo-v2.6-pro`, `mimo-v2.6-pro-ultraspeed`, ...)
 
 The GPT-5.6-and-later family has an additional OpenAI/Azure-context check, so a string containing a qualifying GPT version (for example `gpt-5.6` or `gpt-6`) does not automatically cause GPT-specific fields to be sent to an unrelated endpoint.
 
-MiMo detection targets exactly Flash and Pro: it excludes `mimo-v2.5`,
-`mimo-v2.5-pro`, and `mimo-v2.6-pro-ultraspeed`.
+MiMo detection keeps the documented V2.6 ids and covers future generations after
+V2.6; it excludes `mimo-v2.5`, `mimo-v2.5-pro`, `mimo-v2`, and undocumented V2.6
+variants such as `mimo-v2.6-flashx`.
 
 Unknown models use the neutral policy.
 
@@ -1047,7 +1069,7 @@ returns the runtime-facing descriptor the hooks consume: the legacy policy
 string plus explicit capability flags.
 
 Only registry entries marked `legacy` enable runtime behavior; documented but
-non-legacy entries (for example `mimo-v2.6-pro-ultraspeed`) and all unknown
+non-legacy aliases (for example `gpt-daybreak-blue-latest`) and all unknown
 models resolve to a neutral runtime. A newer or unknown model therefore never
 inherits a current model's mutation unless the registry explicitly registers it.
 The GPT family is a documented exception in the sense that its boundary is
@@ -1282,7 +1304,7 @@ release, use:
 ```json
 {
   "plugin": [
-    "opencode-cache-engine@0.4.1"
+    "opencode-cache-engine@0.4.5"
   ]
 }
 ```
@@ -1385,14 +1407,14 @@ xiaomi/mimo-v2.6-flash
 xiaomi/mimo-v2.6-pro
 ```
 
-`mimo-v2.5`, `mimo-v2.5-pro`, and `mimo-v2.6-pro-ultraspeed` are intentionally
-not matched.
+`mimo-v2.5`, `mimo-v2.5-pro`, `mimo-v2`, and undocumented V2.6 variants such as
+`mimo-v2.6-flashx` are intentionally not matched.
 
 ---
 
 ## OpenRouter affinity header is not added
 
-CacheEngine adds its `x-session-id` only for a detected MiMo-V2.6 or GLM-5.3
+CacheEngine adds its `x-session-id` only for a detected MiMo V2.6-and-later or GLM-5.3-and-later
 request when the actual OpenCode `providerID` is exactly `openrouter`. A direct
 provider route or missing provider identity is bypassed. If a case-insensitive
 `x-session-id` is already present in model or plugin headers, it is preserved
@@ -1423,10 +1445,10 @@ A telemetry failure is intentionally swallowed so it does not break model execut
 The current implementation is intentionally conservative:
 
 ```text
-DeepSeek  -> preserve and measure
-GPT-5.6   -> documented cache key/options; user/harness controls the 272K pricing boundary
-GLM-5.3   -> preserve-content <env> relocation + OpenRouter affinity header
-MiMo-V2.6 -> preserve-content <env> relocation + OpenRouter affinity header
+DeepSeek    -> preserve and measure
+GPT-5.6+    -> documented cache key/options; user/harness controls the 272K pricing boundary
+GLM-5.3+    -> family baseline; GLM-5.3 only: preserve-content <env> relocation + OpenRouter affinity header
+MiMo V2.6+  -> family baseline; V2.6 Flash/Pro only: preserve-content <env> relocation (+ OpenRouter affinity header)
 ```
 
 That separation is the core design of the project.

package/docs/cache-policy-inventory.md

@@ -75,10 +75,10 @@ Detection consequences worth stating explicitly:
   matched → neutral. **[O]**
 - `deepseek-v5` (or any future `*deepseek*` id) matches the passive DeepSeek
   branch because the regex is a bare substring test. **[O]**
-- `mimo-v2.6-pro-ultraspeed` is **not** matched: the `(?![\w-])` lookahead fails
-  on the following `-`. Test `MiMo V2.5 and Pro-UltraSpeed do NOT match MiMo policy`
-  confirms this. **[O]**
-- `mimo-v2.5*` and `glm-5.2`/`glm-4.x` are neutral. **[O]**
+- `mimo-v2.6-pro-ultraspeed` is matched by its own explicit entry since v0.4.5
+  (MiMo family baseline, no `<env>` overlay). **[O]**
+- Undocumented V2.6 variants such as `mimo-v2.6-flashx` remain neutral, and
+  `mimo-v2.5*` and `glm-5.2`/`glm-4.x` are neutral. **[O]**
 
 ---
 
@@ -288,12 +288,12 @@ made here); **hold** = do not inherit without first-party evidence.
 | DeepSeek | future-looking V4+ identifiers: `deepseek-v4.1`, `deepseek-v4`, `deepseek-v5` | Not documented as request ids (`deepseek-v4.1`/`deepseek-v4` invalid or version-string only) | Passive via the V4-and-later family predicate or the safe creator fallback (v0.4.3); no mutation | keep passive; treat as unknown-friendly | none | Medium (detection) / Low (future ids) | DeepSeek *Models & Pricing*; *Chat Completions API* | 2026-09-27 |
 | DeepSeek | pre-V4 negative controls: `deepseek-chat`, `deepseek-reasoner` | Retired names (retired 2026-07-24); no separate V4+ cache policy claimed | Passive | hold | n/a | High | DeepSeek *Change Log*; *news260424* | 2026-09-26 |
 | Z.AI | GLM 5.3: `glm-5.3`, `glm-5.3-flash`, `glm-5.3-flashx` | Implicit automatic caching; `cached_tokens`; stable-prompt-first guidance; no documented min/TTL/key | GLM policy: `<env>` relocation; OpenRouter `x-session-id` | keep (env relocation is an exact overlay, not a Z.AI control) | env relocation is the overlay; keep scoped to GLM-5.3 | Medium | Z.AI *Context Caching*; *Chat Completion*; *Pricing* | 2026-09-26 |
-| Z.AI | current later GLM generations (documented): none newer than 5.3; newest below is `glm-5.2`/`glm-5.1`/`glm-5`/`glm-4.7` | Same implicit mechanism documented service-wide; cached-input price per model | neutral (only `glm-5.3` matched) | **hold** — no doc says 5.3 overlay extends upward; none newer documented | n/a | High (no later gens documented) | Z.AI *New Released*; *Pricing* | 2026-09-26 |
+| Z.AI | current later GLM generations (documented): none newer than 5.3; newest below is `glm-5.2`/`glm-5.1`/`glm-5`/`glm-4.7` | Same implicit mechanism documented service-wide; cached-input price per model | GLM-5.3 family baseline via the 5.3-and-later boundary; **no** `<env>` overlay (v0.4.4) | **baseline only** — overlay stays 5.3-explicit | none documented | High (no later gens documented) | Z.AI *New Released*; *Pricing* | 2026-09-27 |
 | Z.AI | 5.2 and earlier negative controls: `glm-5.2`, `glm-5.1`, `glm-5`, `glm-4.7`, `glm-4.6`, `glm-4.5`, `glm-4-32b-*` | Cacheable (except `glm-4-32b-0414-128k`), different cached-input pricing; no cache-semantics difference documented | neutral | hold | n/a | High | Z.AI *Pricing*; *Chat Completion* (enum) | 2026-09-26 |
 | Xiaomi | MiMo V2.6 Flash: `mimo-v2.6-flash` (OR `xiaomi/mimo-v2.6-flash`) | Provider-managed implicit caching; `cached_tokens`; no documented min/TTL/key/prefix rules | MiMo policy: `<env>` relocation; provider-change telemetry; OpenRouter `x-session-id` | keep scoped as exact-model overlay | env relocation unsupported by docs → treat as overlay | Medium | MiMo *Models*; *Pricing*; *openai-api* | 2026-09-26 |
 | Xiaomi | MiMo V2.6 Pro: `mimo-v2.6-pro` (OR `xiaomi/mimo-v2.6-pro`) | Same documented per-model implicit caching; per-model pricing | MiMo policy (same as Flash) | keep | none documented | Medium | MiMo *Models*; *Pricing* | 2026-09-26 |
-| Xiaomi | MiMo V2.6 Pro UltraSpeed: `mimo-v2.6-pro-ultraspeed` (OR `xiaomi/mimo-v2.6-pro-ultraspeed`) | Documented as a Pro **mode**, same V2.6 series; "Context Caching" listed; no documented mechanism difference from Pro | **Not matched** → neutral (`(?![\w-])` blocks it) | **extend OR document as exception** — docs treat it as the same V2.6 series, but no doc proves identical cache controls | explicit decision needed; current behavior is a de-facto exact-model exception | Medium | MiMo *news/latest/v2-6*; *Models*; *Pricing* | 2026-09-26 |
-| Xiaomi | later MiMo generations: none documented beyond V2.6; V2.5 deprecates 2026-10-21 | Not documented | neutral | hold (no docs) | n/a | High (no later gens documented) | MiMo *Models* | 2026-09-26 |
+| Xiaomi | MiMo V2.6 Pro UltraSpeed: `mimo-v2.6-pro-ultraspeed` (OR `xiaomi/mimo-v2.6-pro-ultraspeed`) | Documented as a Pro **mode**, same V2.6 series; "Context Caching" listed; no documented mechanism difference from Pro | MiMo family baseline only (cached-token telemetry, provider-change/prefix diagnostics, OpenRouter affinity); **no** `<env>` overlay (v0.4.5) | family baseline; overlay stays Flash/Pro-explicit | none — baseline only | Medium | MiMo *news/latest/v2-6*; *Models*; *Pricing* | 2026-09-27 |
+| Xiaomi | later MiMo generations: none documented beyond V2.6; V2.5 deprecates 2026-10-21 | Not documented; no inheritance rule | Passive family baseline for any future >V2.6 id (v0.4.5); no `<env>` overlay | keep passive, baseline only | none | High (no later gens documented) | MiMo *Models*; *updates/model* | 2026-09-27 |
 | Xiaomi | V2.5 negative control: `mimo-v2.5`, `mimo-v2.5-pro` | Documented as deprecated 2026-10-21; cache capability listed; no V2.6 policy inheritance claimed | neutral | hold | n/a | High | MiMo *Models*; *news/latest/v2-6* | 2026-09-26 |
 
 ---
@@ -323,9 +323,8 @@ a newer model inherits an older policy. They must not be resolved by guessing.
    undocumented.
 9. **Xiaomi V2.6 Pro UltraSpeed cache mechanism.** Documented as the same V2.6
    series and a Pro mode, but no first-party statement proves identical cache
-   controls to Pro/Flash. This is the one place where CacheEngine's current
-   non-match (`mimo-v2.6-pro-ultraspeed` → neutral) is a real scope decision
-   rather than a documented fact.
+   controls to Pro/Flash. v0.4.5 resolves this as a CacheEngine scope decision:
+   UltraSpeed gets the family **baseline only** (no `<env>` overlay).
 10. **MiMo on OpenRouter.** OpenRouter's Prompt Caching page documents no
     MiMo-specific cache section; only generic sticky-routing behavior applies.
 11. **OpenAI Chat Completions usage-field naming** (`prompt_tokens_details` vs
@@ -384,3 +383,48 @@ a newer model inherits an older policy. They must not be resolved by guessing.
 - Evidence caveat: first-party pages conflict on whether `deepseek-v4-pro` still
   routes as a distinct model in late 2026; this does not affect the passive
   policy, which carries no mutation either way.
+
+### Follow-up: v0.4.4 GLM-5.3-and-later baseline vs GLM-5.3 overlay
+
+- Z.AI docs were re-verified on **2026-09-27**. GLM-5.3 is the newest documented
+  text generation (no `glm-5.4`/`glm-6`); GLM-5.3 is explicitly "the same base
+  model as GLM-5.2" with post-training differences. Caching is implicit with no
+  `cache_control`/`prompt_cache_key`/breakpoint/TTL field; `cached_tokens` is the
+  reported field. Z.AI publishes **no** generational-inheritance rule.
+- The `<env>` relocation has **no first-party basis** — it is a CacheEngine
+  implementation overlay. v0.4.4 therefore separates the concepts:
+  - **Family baseline** (`zai.glm-5.3-plus`, predicate `glm >= 5.3`): implicit
+    caching plus the non-mutating GLM diagnostics/transport (thinking-integrity
+    telemetry, GLM cache ratio, provider-change telemetry, OpenRouter
+    `x-session-id`). No prompt rewrite.
+  - **GLM-5.3 overlay**: the `<env>` relocation stays registered only on the
+    GLM-5.3 family entry, so a later GLM does **not** inherit it.
+- GLM-5.2 and earlier remain neutral. No new GLM cache-control field is added,
+  and GLM-5.3 behavior is unchanged.
+- Evidence caveat: because no later GLM generation is documented, the
+  "GLM-5.3 and later" boundary is a CacheEngine inference about a passive
+  baseline, not a Z.AI contract; it is safe because it introduces no mutation.
+
+### Follow-up: v0.4.5 MiMo V2.6+ baseline vs the validated MiMo overlay
+
+- Xiaomi MiMo docs were re-verified on **2026-09-27**. The V2.6 series is Pro +
+  Flash, with UltraSpeed documented as a Pro **mode** exposed as
+  `mimo-v2.6-pro-ultraspeed`. No generation newer than V2.6 is documented, and
+  no generational-inheritance rule is published. Caching is provider-managed
+  implicit with no cache-control field in any compat schema; usage fields are
+  per-protocol `cached_tokens`/`cache_read_input_tokens`. V2.5 deprecates
+  2026-10-21.
+- The `<env>` relocation has **no first-party basis** — it is a CacheEngine
+  overlay. v0.4.5 therefore separates:
+  - **Family baseline** (documented V2.6 Flash/Pro, the UltraSpeed mode entry,
+    and any future >V2.6 id via `isMimoAfterV26`): cached-token telemetry,
+    provider-change/prefix diagnostics, OpenRouter `x-session-id`. No prompt
+    rewrite.
+  - **Validated overlay**: the `<env>` relocation stays registered only on the
+    V2.6 Flash/Pro entry.
+- UltraSpeed moves from neutral to the family baseline (no overlay). V2.5 and
+  earlier stay neutral; undocumented V2.6 variants such as `mimo-v2.6-flashx`
+  also stay neutral (narrow detection preserved).
+- No MiMo cache-control field is invented. Evidence caveat: since no >V2.6
+  generation is documented, future coverage is a CacheEngine inference about a
+  passive baseline (safe: no mutation), not a Xiaomi contract.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "opencode-cache-engine",
-  "version": "0.4.3",
+  "version": "0.4.5",
   "private": false,
   "description": "Provider-aware prompt-cache optimization and telemetry for OpenCode",
   "keywords": [

package/src/cache-policy-core.mjs

@@ -102,6 +102,49 @@ export function isDeepseekV4OrLater(slug) {
   return false
 }
 
+// GLM-5.3-and-later family baseline (docs/cache-policy-inventory.md §3; Z.AI
+// docs re-verified 2026-09-27). Z.AI publishes no generational-inheritance rule
+// and no explicit cache control, so this is a CacheEngine inference about a
+// passive/implicit baseline; it is safe because it grants no mutation. The
+// GLM-5.3 `<env>` relocation is a separate, explicit overlay and is NOT granted
+// by this predicate. GLM-5.2 and earlier stay outside this family.
+export function isGlm53OrLater(slug) {
+  const text = String(slug ?? "").toLowerCase()
+  const re = /glm-(\d{1,3})(?:\.(\d))?(?![\d.])/g
+  let m
+  while ((m = re.exec(text)) !== null) {
+    const major = Number(m[1])
+    const minor = m[2] === undefined ? 0 : Number(m[2])
+    if (major > 5) return true
+    if (major === 5 && minor >= 3) return true
+  }
+  return false
+}
+
+// MiMo generations strictly newer than V2.6 (docs/cache-policy-inventory.md §4;
+// Xiaomi docs re-verified 2026-09-27). The documented V2.6 ids (Flash/Pro and
+// the Pro UltraSpeed mode) are matched by their explicit entries; this predicate
+// covers a future V2.7+/V3+ generation that CacheEngine has never seen, granting
+// only the passive baseline (telemetry + transport) and never the `<env>`
+// relocation overlay, so it is safe by construction. V2.5 and earlier stay
+// neutral.
+//
+// It intentionally does NOT match the existing V2.6 literal, so undocumented
+// V2.6 variants (for example `mimo-v2.6-flashx`) stay neutral per the narrow
+// MiMo detection rule.
+export function isMimoAfterV26(slug) {
+  const text = String(slug ?? "").toLowerCase()
+  const re = /mimo-v(\d+)(?:\.(\d+))?(?![\d.])/g
+  let m
+  while ((m = re.exec(text)) !== null) {
+    const major = Number(m[1])
+    const minor = m[2] === undefined ? 0 : Number(m[2])
+    if (major > 2) return true
+    if (major === 2 && minor > 6) return true
+  }
+  return false
+}
+
 // Candidate ids for exact/alias lookup. Includes the raw apiID/modelID, the
 // lower-cased forms, and a single stripped transport/vendor prefix
 // (e.g. "openai/gpt-5.6-luna" -> "gpt-5.6-luna", "xiaomi/mimo-v2.6-flash" ->
@@ -344,6 +387,29 @@ export const POLICY_REGISTRY = [
     }),
     inventoryRef: "§3 Z.AI GLM",
   },
+  {
+    // v0.4.4: "GLM-5.3 and later" family baseline. A future 5.3+ model inherits
+    // the implicit-cache baseline and its non-mutating diagnostics/transport,
+    // but NOT the GLM-5.3-specific `<env>` relocation overlay (`overlays: []`
+    // and no `envRelocation` capability). GLM-5.2 and earlier stay neutral.
+    id: "zai.glm-5.3-plus",
+    creator: "z.ai",
+    family: "glm-5.3",
+    kind: "family",
+    predicate: isGlm53OrLater,
+    baseline: "zai.implicit-cache",
+    overlays: [],
+    legacy: true,
+    runtime: rt("glm53", {
+      thinkingIntegrity: true,
+      cacheRatio: "glm",
+      providerChange: "glm",
+      openRouterAffinity: true,
+    }),
+    boundary: "GLM-5.3 and later",
+    note: "Z.AI publishes no generational-inheritance rule and no explicit cache control. The baseline is implicit caching; the `<env>` relocation is a GLM-5.3-only CacheEngine overlay and is intentionally not inherited. No cache-control field is invented.",
+    inventoryRef: "§3 Z.AI GLM",
+  },
   {
     id: "xiaomi.mimo-v2.6",
     creator: "xiaomi",
@@ -367,14 +433,42 @@ export const POLICY_REGISTRY = [
     id: "xiaomi.mimo-v2.6-pro-ultraspeed",
     creator: "xiaomi",
     family: "mimo-v2.6",
-    kind: "exact",
+    kind: "family",
+    pattern: /mimo-v2\.6-pro-ultraspeed/i,
     exactIds: ["mimo-v2.6-pro-ultraspeed"],
     baseline: "xiaomi.implicit-cache",
     overlays: [],
-    legacy: false,
-    runtime: rt("neutral"),
-    policyStatus: "documented-series-member-without-registered-overlay",
-    note: "Documented as a Pro mode in the same V2.6 series, but the inventory does not establish identical cache controls and CacheEngine registers no overlay for it.",
+    legacy: true,
+    runtime: rt("mimo26", {
+      cacheRatio: "mimo",
+      providerChange: "mimo",
+      prefixDiagnostics: true,
+      openRouterAffinity: true,
+    }),
+    policyStatus: "documented-series-member-baseline-only",
+    note: "v0.4.5: documented as a Pro mode in the same V2.6 series. It gets the MiMo family baseline (cached-token telemetry, provider-change/prefix diagnostics, OpenRouter affinity) but NOT the validated `<env>` relocation overlay.",
+    inventoryRef: "§4 Xiaomi MiMo",
+  },
+  {
+    // v0.4.5: "MiMo V2.6 and later" family baseline. A future 2.6+ model gets
+    // the passive baseline and its non-mutating diagnostics/transport, but never
+    // the `<env>` overlay, which stays explicit to the validated Flash/Pro entry.
+    id: "xiaomi.mimo-v2.6-plus",
+    creator: "xiaomi",
+    family: "mimo-v2.6",
+    kind: "family",
+    predicate: isMimoAfterV26,
+    baseline: "xiaomi.implicit-cache",
+    overlays: [],
+    legacy: true,
+    runtime: rt("mimo26", {
+      cacheRatio: "mimo",
+      providerChange: "mimo",
+      prefixDiagnostics: true,
+      openRouterAffinity: true,
+    }),
+    boundary: "MiMo V2.6 and later",
+    note: "Xiaomi documents implicit caching with no cache-control field and no generational-inheritance rule; the `<env>` relocation is a CacheEngine overlay with no first-party basis and is not inherited.",
     inventoryRef: "§4 Xiaomi MiMo",
   },
   {

package/test/cache-engine.test.mjs

@@ -54,7 +54,9 @@ import {
   OVERLAYS,
   POLICY_REGISTRY,
   isDeepseekV4OrLater,
+  isGlm53OrLater,
   isGpt56OrLater,
+  isMimoAfterV26,
   resolveLegacyFamily,
   resolvePolicy,
   resolveRuntimePolicy,
@@ -324,11 +326,13 @@ test("MiMo V2.6 Flash/Pro match MiMo policy (openrouter + direct)", () => {
   )
 })
 
-test("MiMo V2.5 and Pro-UltraSpeed do NOT match MiMo policy", () => {
+test("MiMo V2.5 stays neutral; V2.6 Pro UltraSpeed now resolves (v0.4.5)", () => {
   assert.equal(detectPolicy(M("openrouter", "xiaomi/mimo-v2.5")), POLICY_NEUTRAL)
   assert.equal(detectPolicy(M("openrouter", "xiaomi/mimo-v2.5-pro")), POLICY_NEUTRAL)
   assert.equal(detectPolicy(M("xiaomi", "mimo-v2.5")), POLICY_NEUTRAL)
-  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-pro-ultraspeed")), POLICY_NEUTRAL)
+  // v0.4.5: UltraSpeed is a documented V2.6 series member (baseline only).
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-pro-ultraspeed")), POLICY_MIMO26)
+  // Undocumented V2.6 variants remain neutral (narrow detection preserved).
   assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-flashx")), POLICY_NEUTRAL)
 })
 
@@ -1447,7 +1451,7 @@ test("resolvePolicy: MiMo V2.6 Flash/Pro carry the overlay; Pro UltraSpeed is an
   assert.equal(ultraspeed.matchType, "exact")
   assert.deepEqual(overlayIds(ultraspeed), [])
   const uEntry = POLICY_REGISTRY.find((e) => e.id === "xiaomi.mimo-v2.6-pro-ultraspeed")
-  assert.equal(uEntry.policyStatus, "documented-series-member-without-registered-overlay")
+  assert.equal(uEntry.policyStatus, "documented-series-member-baseline-only")
 
   const v25 = resolvePolicy(M("xiaomi", "mimo-v2.5"))
   assert.equal(v25.family, "neutral")
@@ -1650,10 +1654,15 @@ async function runPolicyMigrationProbe() {
       { name: "deepseek-gateway", model: { providerID: "acme-gateway", id: "my-deepseek-mirror", api: { id: "my-deepseek-mirror" } }, expect: { policy: "deepseek", env: false, gpt: false, header: false } },
       { name: "glm-5.3-direct", model: { providerID: "zai", id: "glm-5.3", api: { id: "glm-5.3" } }, expect: { policy: "glm53", env: true, gpt: false, header: false } },
       { name: "glm-5.3-openrouter", model: { providerID: "openrouter", id: "z-ai/glm-5.3-flash", api: { id: "z-ai/glm-5.3-flash" } }, expect: { policy: "glm53", env: true, gpt: false, header: true } },
+      { name: "glm-5.4-direct", model: { providerID: "zai", id: "glm-5.4", api: { id: "glm-5.4" } }, expect: { policy: "glm53", env: false, gpt: false, header: false } },
+      { name: "glm-5.4-openrouter", model: { providerID: "openrouter", id: "z-ai/glm-5.4", api: { id: "z-ai/glm-5.4" } }, expect: { policy: "glm53", env: false, gpt: false, header: true } },
       { name: "glm-5.2", model: { providerID: "zai", id: "glm-5.2", api: { id: "glm-5.2" } }, expect: { policy: "neutral", env: false, gpt: false, header: false } },
       { name: "mimo-v2.6-flash-direct", model: { providerID: "xiaomi", id: "mimo-v2.6-flash", api: { id: "mimo-v2.6-flash" } }, expect: { policy: "mimo26", env: true, gpt: false, header: false } },
       { name: "mimo-v2.6-pro-openrouter", model: { providerID: "openrouter", id: "xiaomi/mimo-v2.6-pro", api: { id: "xiaomi/mimo-v2.6-pro" } }, expect: { policy: "mimo26", env: true, gpt: false, header: true } },
-      { name: "mimo-v2.6-pro-ultraspeed", model: { providerID: "xiaomi", id: "mimo-v2.6-pro-ultraspeed", api: { id: "mimo-v2.6-pro-ultraspeed" } }, expect: { policy: "neutral", env: false, gpt: false, header: false } },
+      { name: "mimo-v2.6-pro-ultraspeed", model: { providerID: "xiaomi", id: "mimo-v2.6-pro-ultraspeed", api: { id: "mimo-v2.6-pro-ultraspeed" } }, expect: { policy: "mimo26", env: false, gpt: false, header: false } },
+      { name: "mimo-v2.7-future", model: { providerID: "xiaomi", id: "mimo-v2.7-flash", api: { id: "mimo-v2.7-flash" } }, expect: { policy: "mimo26", env: false, gpt: false, header: false } },
+      { name: "mimo-v2.7-openrouter", model: { providerID: "openrouter", id: "xiaomi/mimo-v2.7-flash", api: { id: "xiaomi/mimo-v2.7-flash" } }, expect: { policy: "mimo26", env: false, gpt: false, header: true } },
+      { name: "mimo-v2.6-pro-ultraspeed-openrouter", model: { providerID: "openrouter", id: "xiaomi/mimo-v2.6-pro-ultraspeed", api: { id: "xiaomi/mimo-v2.6-pro-ultraspeed" } }, expect: { policy: "mimo26", env: false, gpt: false, header: true } },
       { name: "mimo-v2.5", model: { providerID: "xiaomi", id: "mimo-v2.5", api: { id: "mimo-v2.5" } }, expect: { policy: "neutral", env: false, gpt: false, header: false } },
       { name: "unknown-provider", model: { providerID: "mystery-provider", id: "xiaomi/mimo-v2.6-flash", api: { id: "xiaomi/mimo-v2.6-flash" } }, expect: { policy: "mimo26", env: true, gpt: false, header: false } },
       { name: "unknown-openrouter", model: { providerID: "openrouter", id: "acme/mystery-9", api: { id: "acme/mystery-9" } }, expect: { policy: "neutral", env: false, gpt: false, header: false } },
@@ -1915,3 +1924,194 @@ test("v0.4.3: DeepSeek never receives OpenRouter affinity or GPT/GLM/MiMo fields
     assert.equal(r.existingHeadersPreserved, true, `${name}: headers preserved`)
   }
 })
+
+// ===========================================================================
+// v0.4.4 GLM-5.3-and-later family baseline vs the GLM-5.3-specific overlay
+//
+// Source: Z.AI docs re-verified 2026-09-27 (docs/cache-policy-inventory.md §3).
+// Z.AI documents implicit caching (no cache control) and publishes no
+// generational-inheritance rule; the `<env>` relocation is a CacheEngine overlay
+// with no first-party basis. GLM-5.2 and earlier stay neutral.
+// ===========================================================================
+
+test("v0.4.4: isGlm53OrLater matches GLM-5.3+ version tokens only", () => {
+  const inFamily = ["glm-5.3", "glm-5.3-flash", "glm-5.3-flashx", "glm-5.4", "glm-5.9", "glm-6", "z-ai/glm-5.3-flash"]
+  for (const id of inFamily) assert.equal(isGlm53OrLater(id), true, `${id} is 5.3+`)
+  const outOfFamily = ["glm-5.2", "glm-5.1", "glm-5", "glm-4.7", "glm-4.6", "glm-4.5", "glm-4.5-air", "glm-4-32b-0414-128k", ""]
+  for (const id of outOfFamily) assert.equal(isGlm53OrLater(id), false, `${id} is pre-5.3`)
+})
+
+test("v0.4.4: GLM-5.3 keeps its overlay; a later GLM gets the baseline only", () => {
+  const g53 = resolvePolicy(M("zai", "glm-5.3"))
+  assert.equal(g53.family, "glm-5.3")
+  assert.equal(baseId(g53), "zai.implicit-cache")
+  assert.deepEqual(overlayIds(g53), ["glm53.env-relocation"])
+
+  const g54 = resolvePolicy(M("zai", "glm-5.4"))
+  assert.equal(g54.creator, "z.ai")
+  assert.equal(g54.family, "glm-5.3")
+  assert.equal(baseId(g54), "zai.implicit-cache") // same family baseline
+  assert.deepEqual(overlayIds(g54), []) // overlay is NOT inherited
+
+  // Runtime capability separation: baseline diagnostics/transport yes, prompt rewrite no.
+  const c53 = resolveRuntimePolicy(M("zai", "glm-5.3"))
+  const c54 = resolveRuntimePolicy(M("zai", "glm-5.4"))
+  assert.equal(c54.policy, "glm53")
+  assert.equal(c54.envRelocation, null)
+  assert.equal(c54.thinkingIntegrity, true)
+  assert.equal(c54.cacheRatio, "glm")
+  assert.equal(c54.providerChange, "glm")
+  assert.equal(c54.openRouterAffinity, true)
+  assert.equal(c53.envRelocation, "glm")
+
+  // Boundary metadata is traceable and the overlay is registered separately.
+  const plus = POLICY_REGISTRY.find((e) => e.id === "zai.glm-5.3-plus")
+  assert.equal(plus.boundary, "GLM-5.3 and later")
+  assert.deepEqual(plus.overlays, [])
+  assert.ok(plus.inventoryRef)
+})
+
+test("v0.4.4: GLM-5.2 and earlier stay neutral", () => {
+  for (const id of ["glm-5.2", "glm-5.1", "glm-5", "glm-4.7", "glm-4.6", "glm-4.5"]) {
+    const r = resolvePolicy(M("zai", id))
+    assert.equal(r.family, "neutral", `${id} neutral`)
+    assert.deepEqual(overlayIds(r), [])
+    assert.equal(resolveRuntimePolicy(M("zai", id)).policy, "neutral")
+  }
+})
+
+test("v0.4.4: legacy detectPolicy follows the GLM-5.3-and-later boundary", () => {
+  assert.equal(detectPolicy(M("zai", "glm-5.3-flash")), POLICY_GLM53)
+  assert.equal(detectPolicy(M("zai", "glm-5.4")), POLICY_GLM53)
+  assert.equal(detectPolicy(M("zai", "glm-5.2")), POLICY_NEUTRAL)
+})
+
+test("v0.4.4: <env> absent leaves system content unchanged", () => {
+  const plain = "Stable instructions only.\nNo environment block here."
+  const r = relocateVolatileEnvBlock(plain)
+  assert.equal(r.changed, false)
+  assert.equal(r.text, plain)
+})
+
+test("v0.4.4: later GLM inherits the baseline but never the <env> rewrite (runtime)", async () => {
+  const { results } = await policyMigrationResults()
+  const g53 = results.find((r) => r.name === "glm-5.3-direct")
+  const g54 = results.find((r) => r.name === "glm-5.4-direct")
+  assert.ok(g53 && g54)
+  // Identical system text with a valid <env> block present in both cases.
+  assert.equal(g53.systemRelocated, true) // 5.3 overlay fires
+  assert.equal(g54.systemRelocated, false) // later GLM does NOT inherit it
+  assert.equal(g54.runtimePolicy, "glm53") // but the family baseline applies
+  assert.equal(g54.gptOptionInjected, false)
+  assert.equal(g54.affinityHeaderAttached, false)
+})
+
+test("v0.4.4: GLM OpenRouter affinity is transport-gated for later GLM too", async () => {
+  const { results } = await policyMigrationResults()
+  const or = results.find((r) => r.name === "glm-5.4-openrouter")
+  const direct = results.find((r) => r.name === "glm-5.4-direct")
+  assert.equal(or.affinityHeaderAttached, true)
+  assert.equal(or.systemRelocated, false)
+  assert.equal(direct.affinityHeaderAttached, false)
+  assert.equal(or.existingHeadersPreserved, true)
+  assert.equal(direct.existingHeadersPreserved, true)
+})
+
+// ===========================================================================
+// v0.4.5 MiMo V2.6+ family baseline vs the validated MiMo overlay
+//
+// Source: Xiaomi MiMo docs re-verified 2026-09-27
+// (docs/cache-policy-inventory.md §4). Caching is provider-managed/implicit with
+// no cache-control field; the `<env>` relocation is a CacheEngine overlay with
+// no first-party basis; V2.5 deprecates 2026-10-21. No V2.7+/inheritance rule is
+// documented, so future coverage is a safe passive baseline only.
+// ===========================================================================
+
+test("v0.4.5: isMimoAfterV26 matches generations strictly after V2.6", () => {
+  const inFamily = ["mimo-v2.7-flash", "mimo-v2.7", "mimo-v2.8-pro", "mimo-v3", "mimo-v3.1-flash", "xiaomi/mimo-v2.7-flash"]
+  for (const id of inFamily) assert.equal(isMimoAfterV26(id), true, `${id} is after V2.6`)
+  const outOfFamily = ["mimo-v2.6-flash", "mimo-v2.6-pro", "mimo-v2.6-pro-ultraspeed", "mimo-v2.6-flashx", "mimo-v2.5", "mimo-v2.5-pro", "mimo-v2-flash", ""]
+  for (const id of outOfFamily) assert.equal(isMimoAfterV26(id), false, `${id} is not after V2.6`)
+})
+
+test("v0.4.5: Flash/Pro keep the overlay; UltraSpeed and later get the baseline only", () => {
+  const flash = resolvePolicy(M("xiaomi", "mimo-v2.6-flash"))
+  assert.equal(flash.family, "mimo-v2.6")
+  assert.equal(baseId(flash), "xiaomi.implicit-cache")
+  assert.deepEqual(overlayIds(flash), ["mimo26.env-relocation"])
+  assert.deepEqual(overlayIds(resolvePolicy(M("xiaomi", "mimo-v2.6-pro"))), ["mimo26.env-relocation"])
+
+  for (const id of ["mimo-v2.6-pro-ultraspeed", "mimo-v2.7-flash"]) {
+    const r = resolvePolicy(M("xiaomi", id))
+    assert.equal(r.family, "mimo-v2.6", `${id}: family`)
+    assert.equal(baseId(r), "xiaomi.implicit-cache", `${id}: same baseline`)
+    assert.deepEqual(overlayIds(r), [], `${id}: no overlay`)
+    const c = resolveRuntimePolicy(M("xiaomi", id))
+    assert.equal(c.policy, "mimo26", `${id}: policy`)
+    assert.equal(c.envRelocation, null, `${id}: no prompt rewrite`)
+    assert.equal(c.cacheRatio, "mimo", `${id}: cached-token ratio`)
+    assert.equal(c.providerChange, "mimo", `${id}: provider diagnostics`)
+    assert.equal(c.openRouterAffinity, true, `${id}: affinity capability`)
+    assert.equal(c.gptCacheMetadata, false, `${id}: no leak`)
+  }
+
+  // Same baseline object family for the overlay and non-overlay members.
+  assert.equal(baseId(resolvePolicy(M("xiaomi", "mimo-v2.7-flash"))), baseId(flash))
+})
+
+test("v0.4.5: MiMo V2.5 and earlier stay neutral", () => {
+  for (const id of ["mimo-v2.5", "mimo-v2.5-pro", "mimo-v2-flash", "mimo-v2"]) {
+    const r = resolvePolicy(M("xiaomi", id))
+    assert.equal(r.family, "neutral", `${id}: neutral`)
+    assert.deepEqual(overlayIds(r), [], `${id}: no overlay`)
+    assert.equal(resolveRuntimePolicy(M("xiaomi", id)).policy, "neutral", `${id}: neutral runtime`)
+  }
+})
+
+test("v0.4.5: legacy detectPolicy follows the MiMo V2.6+ boundary", () => {
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-flash")), POLICY_MIMO26)
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-pro-ultraspeed")), POLICY_MIMO26)
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.7-flash")), POLICY_MIMO26)
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.5")), POLICY_NEUTRAL)
+  assert.equal(detectPolicy(M("xiaomi", "mimo-v2.6-flashx")), POLICY_NEUTRAL)
+})
+
+test("v0.4.5: <env> absent is a no-op for MiMo transforms", () => {
+  const plain = "Stable MiMo instructions.\nNo environment block."
+  const r = relocateVolatileEnvBlock(plain)
+  assert.equal(r.changed, false)
+  assert.equal(r.text, plain)
+})
+
+test("v0.4.5: later MiMo inherits the baseline but never the <env> overlay (runtime)", async () => {
+  const { results } = await policyMigrationResults()
+  const flash = results.find((r) => r.name === "mimo-v2.6-flash-direct")
+  const ultra = results.find((r) => r.name === "mimo-v2.6-pro-ultraspeed")
+  const future = results.find((r) => r.name === "mimo-v2.7-future")
+  assert.ok(flash && ultra && future)
+  // Identical system text with a valid <env> block present in every case.
+  assert.equal(flash.systemRelocated, true) // validated Flash overlay fires
+  assert.equal(ultra.systemRelocated, false) // UltraSpeed: baseline only
+  assert.equal(future.systemRelocated, false) // future model: baseline only
+  assert.equal(ultra.runtimePolicy, "mimo26")
+  assert.equal(future.runtimePolicy, "mimo26")
+})
+
+test("v0.4.5: MiMo OpenRouter affinity stays transport-gated", async () => {
+  const { results } = await policyMigrationResults()
+  const orFuture = results.find((r) => r.name === "mimo-v2.7-openrouter")
+  const orUltra = results.find((r) => r.name === "mimo-v2.6-pro-ultraspeed-openrouter")
+  const direct = results.find((r) => r.name === "mimo-v2.7-future")
+  assert.equal(orFuture.affinityHeaderAttached, true)
+  assert.equal(orUltra.affinityHeaderAttached, true)
+  assert.equal(direct.affinityHeaderAttached, false)
+  assert.equal(orFuture.systemRelocated, false) // affinity is transport, not the overlay
+})
+
+test("v0.4.5: MiMo baseline invents no cache fields and keeps cached-token telemetry", () => {
+  const c = resolveRuntimePolicy(M("xiaomi", "mimo-v2.7-flash"))
+  assert.equal(c.gptCacheMetadata, false)
+  assert.equal(c.envRelocation, null)
+  assert.equal(c.cacheRatio, "mimo") // cachedTokens / promptTokens
+  assert.equal(mimoHitRate(75, 100), 75)
+})