feat(localization): add StringTable discovery and validation

CHANGELOG.md

@@ -23,6 +23,7 @@ Two further internal changes, neither of them a release gate:
 
 ### Added
 
+- **Read-only localization preflight is now a first-class namespace.** `localization` adds four cohesive actions from `MonolithConfig`: `list_cultures`, `list_string_tables`, `get_string_table`, and `validate_string_table`. Discovery is Asset Registry-backed and stably paginated; entry readback uses an exclusive key cursor plus independent metadata and text budgets; validation reports deterministic bounded issues and refuses to call an incomplete scan valid. The dispatcher advertises `readOnlyHint=true` and `idempotentHint=true`, registers even when config authoring is disabled, and exposes no create, set, import, export, transaction, dirty, or save path.
 - **Animation layers can be authored without an interface asset.** New `animation add_anim_layer_graph` creates an ABP-native animation layer: a `UAnimationGraph` on the animation graph schema, in the Animation Blueprint's own function graphs — exactly what the editor's My Blueprint → **+** → Animation Layer button produces. Because the layer belongs to the ABP rather than to a shared `UAnimLayerInterface`, variant ABPs no longer have to agree on a layer signature just to have a layer. The schema is the part that matters: it is what makes the animation compiler emit a real `FAnimBlueprintFunction` for the graph, which is why `blueprint add_function` could not stand in — it produces an inert K2 graph. The Output Pose root node is created automatically, since a layer graph without one fails to compile. Optional `input_poses` declares input pose pins by name (`["InPose"]` or `[{"name": "InPose"}]`, capped at 16). The action refuses a duplicate graph name outright rather than renaming the incumbent graph the way the underlying engine call does, and also refuses the reserved name `AnimGraph`, child Animation Blueprints, macro libraries, and interface Blueprints.
 - **`monolith_discover` can search every namespace at once.** The filter predicate lived inside the namespace branch, and the no-namespace branch parsed no `filter`, `offset`, `limit` or `detail` at all — so `monolith_discover(filter="set_property")` with no namespace was silently ignored, handed back the plain namespace inventory, and finding a capability by partial name meant walking every namespace by hand. A non-empty `filter` with no `namespace` now searches the whole registry in registry order, tagging each row with its owning namespace and reporting `matched_namespaces` before pagination, so "which namespace owns this" stays answerable even when the rows are capped. The per-namespace path is naturally bounded at roughly 150 actions and this one is not, so an absent `limit` caps at 50 while an explicit `limit: 0` still means all, matching the existing contract. Mirrored in the offline CLI and `monolith_offline.py`. Reported and prototyped by **@kunkunGames** (#112).
 - **`risk get_mining_status` says why a risk query came back empty.** New read-only action reporting which repositories were mined, which candidates were rejected and for what reason, whether mining ran this session and what it reported, the row counts of the four risk tables, and the stored-versus-current version and configuration stamps. A missing `EngineSource.db` is not an error for it: the settings and a live root resolution are still reported, which is exactly the situation a caller reaches for it in. The four git-backed `risk` actions additionally attach a `diagnostics` block **when they return nothing**, carrying the same scanned and skipped lists plus, when nothing at all was mined, a one-line hint naming the setting to change. `list_conditional_gates` does not carry it — its data comes from the source sweep rather than from git, so a repository hint there would be wrong. The indexer's own summary moves from `Verbose` to `Warning` when it mines nothing.

Docs/API_REFERENCE.md

@@ -1,6 +1,6 @@
 # Monolith API Reference
 
-**Version:** v0.22.0 · **Last updated:** 2026-08-01
+**Version:** v0.22.0 · **Last updated:** 2026-08-04
 
 **In-tree action total is approximate: ~1,400+ actions across 25+ in-tree namespaces** (public, in-tree only; all active by default, plus 45 experimental town-gen actions that register only when `bEnableProceduralTownGen=true`). The surface is too large to track to the unit — **query `monolith_discover()` (its `total_actions` field) for the exact live figure.** The `ui` namespace re-exports 4 GAS UI binding actions as aliases. v0.19.0 adds an LLM C++ authoring ergonomics pack (`source`, 8 actions + `editor.get_build_errors` fix hints), live-PIE introspection + driving and stat-group readout (`editor`), anim-node binding read/write and time-series PIE sampling (`animation`), a Blueprint variable census + contract reconciliation (`blueprint`), and T3D asset-text export (`project`); plus two first-launch fixes (issue #70) and a ~40% smaller `tools/list` manifest. The `monolith_*` meta-tools (`discover`, `status`, `update`, `reindex`, `guide`) plus the `bulk_fill_query` and `describe_query` framework dispatchers round out the MCP tool count. This total EXCLUDES sibling-plugin actions — they ship in their own repos and are never in the public release zip.
 
@@ -24,7 +24,8 @@ The per-namespace numbers in the Table of Contents and body sections below are k
 | [animation](#animation) | 145 | Curves, bone tracks, sync markers, root motion, compression, blend spaces (incl. baking + interpolation control), ABPs (incl. an AnimGraph-authoring pack — additive/slot/cached-pose/blend (by int + by enum)/sync/layered-blend/Control Rig/linked-layer/conduit nodes + output wiring — ABP-native animation layer graphs, custom anim-graph nodes + state-machine teardown + compound expression transition rules), montages, skeletons, PoseSearch, IKRig, Control Rig |
 | [niagara](#niagara) | 119 | Niagara VFX (emitters, modules, params, renderers, HLSL, dynamic inputs, event handlers, sim stages, effect types, event-aware summaries + validate_system event-chain reasoning, temporal-control composite writers + read aggregators, stateless-emitter factory) |
 | [editor](#editor) | 29 | Live Coding builds, compile output capture, editor logs, scene capture, texture import, map creation, module status, automation test list/run, Python escape-hatch, persistent-level swap |
-| [config](#config) | 6 | INI config inspection and search |
+| [config](#config) | 7 | INI config inspection, search, and developer-setting authoring |
+| [localization](#localization) | 4 | Read-only culture discovery plus bounded StringTable inspection and validation |
 | [project](#project) | 7 | Project-wide asset index (SQLite + FTS5) |
 | [source](#source) | 11 | Unreal Engine C++ source code navigation |
 | [mesh](#mesh) | 194 | Mesh inspection, scene manipulation, spatial queries, blockout, GeometryScript, procedural geo, lighting, audio, performance, mesh import (incl. skeletal + animation). +45 town gen registers only with `bEnableProceduralTownGen=true` (experimental, not in the public count) |
@@ -43,7 +44,7 @@ The per-namespace numbers in the Table of Contents and body sections below are k
 | [network](#network) | 4 | **New v0.17.0.** Reflection Intelligence — UE 5.7 replication inspection (replicated classes, RPCs, OnRep handlers, unbalanced-OnRep audit) |
 | [pipeline](#pipeline) | 2 | **New v0.17.0.** Reflection Intelligence — read-only composer actions (`pr_review`, `release_readiness`) |
 | [reflect](#reflect) | 1 | **New v0.19.0.** Reflection Intelligence — index maintenance (`rebuild_reflection_index`, project-only force-rebuild of the RI reflection tables; WRITE/maintenance) |
-| **In-tree subtotal** | **1406** | (all default-active; +45 experimental town gen → 1451 when registered) |
+| **In-tree subtotal** | **1410** | (all default-active; +45 experimental town gen → 1455 when registered) |
 | [Sibling plugins](#sibling-plugins) | varies | Separate plugins, separate distribution |
 
 ---
@@ -654,7 +655,7 @@ Close the current persistent level (without saving) and load the specified level
 
 ## config
 
-INI config file inspection and search. **6 actions.** Read-only.
+INI config file inspection, search, and developer-setting authoring. **7 actions.** The first six actions below are read-only; `set_developer_setting` is an editor-only write.
 
 ### `config.resolve_setting`
 
@@ -712,6 +713,40 @@ List all config files with their hierarchy level.
 |-----------|------|----------|-------------|
 | `category` | string | optional | Filter to a specific category |
 
+### `config.set_developer_setting`
+
+Set one reflected property on a `UDeveloperSettings` class default object. This editor-only action accepts a short or fully qualified class name, parses the requested value through the property's Unreal text importer, and optionally persists it through `SaveConfig`.
+
+| Parameter | Type | Required | Description |
+|-----------|------|----------|-------------|
+| `class` | string | **required** | `UDeveloperSettings` class name or path |
+| `property` | string | **required** | Reflected property name |
+| `value` | string | **required** | Unreal text representation accepted by the property |
+| `save_config` | boolean | optional | Persist to the class config file. Default: `false` |
+
+---
+
+## localization
+
+Read-only culture discovery plus bounded StringTable discovery, readback, and validation. The namespace is registered by `MonolithConfig` independently of the config-authoring toggle and advertises read-only and idempotent dispatcher hints.
+
+All asset inputs require canonical mounted package paths or matching top-level object paths. Pagination, cursor continuation, metadata budgets, text truncation, validation scan limits, and issue pagination are explicit so a partial result is never presented as complete.
+
+| Action | Parameters | Returns |
+|---|---|---|
+| `list_cultures` | `culture_names?` (max 256), `include_derived=true`, `offset=0`, `limit=100` (1–500) | Current culture/language/locale plus a stable culture page and unresolved explicit names |
+| `list_string_tables` | `path=/Game`, `offset=0`, `limit=200` (1–1000), `include_details=false` | Stable Asset Registry page; optional details load only returned assets and include namespace/entry count |
+| `get_string_table` | `asset_path`, `after_key?`, `entry_limit=200` (1–1000), `include_metadata=false`, `metadata_limit=512` (0–4096), `text_limit=4096` (1–65536) | Stable exclusive-key page with independent entry, metadata, and text completeness fields |
+| `validate_string_table` | `asset_path`, `scan_limit=4096` (1–10000), `issue_offset=0`, `issue_limit=200` (1–1000) | Empty-table and scan-cutoff errors; key-edge-whitespace and empty-source warnings; deterministic issue page plus `valid` and `complete` |
+
+For `get_string_table`, continue with `next_after_key` only when `has_more_entries=true`. `all_entries_covered` is true only for a first page that reaches the end; `metadata_complete` independently reports whether the shared metadata budget covered every metadata row on the returned entries. Global `complete` requires both conditions.
+
+For validation, `valid=true` requires `complete=true` and zero errors. A scan cutoff is itself an error. Issue pagination does not change the total error/warning counts or validation verdict.
+
+The existing `blueprint` StringTable actions remain the authoring surface. The `localization` namespace intentionally contains no create, set, remove, import, export, transaction, dirty, or save action.
+
+See `Plugins/Monolith/Skills/unreal-localization/SKILL.md` for routing and bounded workflow examples.
+
 ---
 
 ## project