Add C++ local-document ingest with immutable chunks

godbrain_core/cpp_memory_store/README.md

@@ -1,4 +1,4 @@
-# C++ Alexandria Memory Store (cut 10)
+# C++ Alexandria Memory Store (cut 11)
 
 Same stdin JSON door as Go `memory-store.exe`. Go under `godbrain_core/memory_store/`
 stays. This tree is the C++ replacement; Start/Heal still launch Go until this
@@ -49,6 +49,10 @@ Cut 10: `rag-eval.exe` offline hybrid fixture (`-corpus` path, default Go
 testdata). `-live` still uses Go. Thresholds match Go (Recall/MRR/nDCG ≥ 0.90,
 citation 1.0, no leakage).
 
+Cut 11: ingest `document` + `chunks` (Local-Document-Adapter). Same gates as Go:
+metadata and chunks together, `content_sha256` of `raw_transcript`, contiguous
+UTF-8 byte ranges, forbidden-secret scan, immutable `chunks` collection.
+
 Start/Heal still launch Go.
 
 ```powershell

godbrain_core/cpp_memory_store/include/godbrain/memory_store/embedding.hpp

@@ -48,6 +48,7 @@ bool embedding_cosine(const std::vector<float>& left, const std::vector<float>&
 bool embedding_embed(
     const EmbeddingRuntime& runtime, const std::string& content, std::vector<float>* vector, std::string* err);
 bool embedding_embed_fake(int dimension, const std::string& content, std::vector<float>* vector, std::string* err);
+bool sha256_hex(const std::string& bytes, std::string* hex, std::string* err);
 
 int run_embedding_self_test();
 

godbrain_core/cpp_memory_store/include/godbrain/memory_store/protocol.hpp

@@ -52,6 +52,30 @@ struct Claim {
     std::vector<std::string> evidence_spans;
 };
 
+struct DocumentMetadata {
+    std::string source_label;
+    std::string display_name;
+    std::string file_sha256;
+    std::string content_sha256;
+    std::string extraction_method;
+    std::vector<std::string> languages;
+    std::string backend;
+    std::string backend_version;
+    int chunk_count = 0;
+    bool has_ocr_confidence = false;
+    double ocr_confidence = 0;
+};
+
+struct SourceChunk {
+    int index = 0;
+    int count = 0;
+    int start_byte = 0;
+    int end_byte = 0;
+    std::string text;
+    bool has_confidence = false;
+    double confidence = 0;
+};
+
 struct SkillExtracted {
     std::string name;
     std::string content;
@@ -77,8 +101,12 @@ struct DistillationPayload {
     std::vector<std::string> opsec_candidates;
     std::vector<SkillExtracted> skills_extracted;
     bool has_document = false;
+    DocumentMetadata document;
+    std::vector<SourceChunk> chunks;
 };
 
+bool validate_document_payload(const DistillationPayload& p, std::string* err);
+
 std::string json_escape(const std::string& s);
 std::string skill_procedure_text(const SkillExtracted& skill);
 std::string skill_stable_id(const std::string& name, const std::string& content);

godbrain_core/cpp_memory_store/src/embedding.cpp

@@ -146,6 +146,8 @@ std::wstring fields_join(const std::wstring& in) {
     return out;
 }
 
+}  // namespace
+
 bool sha256_hex(const std::string& bytes, std::string* hex, std::string* err) {
     BCRYPT_ALG_HANDLE alg = nullptr;
     NTSTATUS st = BCryptOpenAlgorithmProvider(&alg, BCRYPT_SHA256_ALGORITHM, nullptr, 0);
@@ -174,6 +176,8 @@ bool sha256_hex(const std::string& bytes, std::string* hex, std::string* err) {
     return true;
 }
 
+namespace {
+
 std::string wide_lower(const std::wstring& w) {
     if (w.empty()) return "";
     std::wstring copy = w;