Delete rag

rag/build_index.py

@@ -1,32 +0,0 @@
-import os
-import faiss
-import pickle
-from sentence_transformers import SentenceTransformer
-
-CORPUS_PATH = "../mcma/micro-cyber-llm/rag/corpus/malware_knowledge.txt"
-OUT_DIR = "../mcma/micro-cyber-llm/rag/vectorstore"
-
-os.makedirs(OUT_DIR, exist_ok=True)
-
-print("[*] Loading embedding model...")
-model = SentenceTransformer("all-MiniLM-L6-v2")
-
-print("[*] Reading corpus...")
-with open(CORPUS_PATH, "r", encoding="utf-8") as f:
-    documents = [line.strip() for line in f if line.strip()]
-
-print(f"[*] Embedding {len(documents)} documents...")
-embeddings = model.encode(documents, show_progress_bar=True)
-
-dim = embeddings.shape[1]
-index = faiss.IndexFlatL2(dim)
-index.add(embeddings)
-
-print("[*] Saving FAISS index...")
-faiss.write_index(index, f"{OUT_DIR}/index.faiss")
-
-print("[*] Saving metadata...")
-with open(f"{OUT_DIR}/meta.pkl", "wb") as f:
-    pickle.dump(documents, f)
-
-print("[✓] RAG index built successfully!")

rag/corpus/malware_knowledge.txt

@@ -1,13 +0,0 @@
-Android malware often abuses READ_SMS to intercept OTP messages.
-Communication with api.telegram.org is commonly used for C2 exfiltration.
-Banking trojans target SMS permissions and overlay attacks.
-APK files requesting SMS and internet permissions are high risk.
-
-Windows malware may use CreateRemoteThread for process injection.
-Suspicious EXE files often drop persistence via registry Run keys.
-C2 traffic over HTTPS to unknown domains is a red flag.
-PowerShell abuse is common in post-exploitation.
-
-MITRE T1406 refers to SMS Control.
-MITRE T1055 refers to Process Injection.
-MITRE T1059 refers to Command and Scripting Interpreter.

rag/ingest.py

@@ -1,14 +0,0 @@
-from sentence_transformers import SentenceTransformer
-import faiss, os, pickle
-
-model = SentenceTransformer("all-MiniLM-L6-v2")
-index = faiss.IndexFlatL2(384)
-docs = []
-
-def ingest(text):
-    emb = model.encode([text])
-    index.add(emb)
-    docs.append(text)
-
-    faiss.write_index(index, "rag/vectorstore/index.faiss")
-    pickle.dump(docs, open("rag/vectorstore/docs.pkl", "wb"))

rag/search.py

@@ -1,20 +0,0 @@
-import os
-import faiss
-import pickle
-from sentence_transformers import SentenceTransformer
-
-BASE_DIR = os.path.dirname(os.path.dirname(__file__))
-VECTOR_DIR = os.path.join(BASE_DIR, "rag", "vectorstore")
-
-index = faiss.read_index(os.path.join(VECTOR_DIR, "index.faiss"))
-
-with open(os.path.join(VECTOR_DIR, "meta.pkl"), "rb") as f:
-    documents = pickle.load(f)
-
-model = SentenceTransformer("sentence-transformers/all-MiniLM-L6-v2")
-
-
-def search_context(query, top_k=3):
-    q_emb = model.encode([query])
-    D, I = index.search(q_emb, top_k)
-    return "\n".join([documents[i] for i in I[0]])

rag/vectorstore/meta.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:043712a56c3aa8d0e33c60438f2a3d2fed218fe29a3e62b04f10f30ce7bf98ff
-size 673