diff --git a/Makefile b/Makefile
index 001b289..d9052e6 100644
--- a/Makefile
+++ b/Makefile
@@ -1,6 +1,6 @@
 .PHONY: linux mails/bpf data/feature-versions.yaml agent/questionaire
 linux:
-	git clone https://github.com/torvalds/linux
+	git clone https://github.com/torvalds/linux --depth=1
 
 mails/bpf:
 	cd mails && git clone --mirror http://lore.kernel.org/bpf/0 bpf/git/0.git
diff --git a/README.md b/README.md
index fca4ee4..a3fabfb 100644
--- a/README.md
+++ b/README.md
@@ -28,8 +28,8 @@ The result can be found in [report_ebpf.md](report_ebpf.md).
 
 - Researcher has proved that LLM can acheive similar level as human when doing summary/survey/questionare task in Market, Chemistry... Such system is also deployed and explored in production for... (See the citation)
 - why not other methods?
-  - domain knowledgeit（cite)
-  - unstruct data （cite）
+  - domain knowledge (cite)
+  - unstructured data（cite）
   - human expert is too expensive and impossible for such task
  
 What question can llm help answer but other cannot?
@@ -55,3 +55,12 @@ What question can llm help answer but other cannot?
 
 ## reference
 
+1. domain knowledge
+   1. [Knowledge solver: Teaching llms to search for domain knowledge from knowledge graphs](https://proceedings.neurips.cc/paper_files/paper/2023/hash/1190733f217404edc8a7f4e15a57f301-Abstract-Datasets_and_Benchmarks.html)
+   2. [Openagi: When llm meets domain experts](https://arxiv.org/abs/2309.03118)
+2. unstructured data
+   1. [Embedding-based Retrieval with LLM for Effective Agriculture Information Extracting from Unstructured Data](https://arxiv.org/abs/2308.03107)
+3. kernel mail
+   1. [How to Communicate when Submitting Patches: An Empirical Study of the Linux Kernel](https://dl.acm.org/doi/abs/10.1145/3359210?casa_token=5CrG9X-8QNgAAAAA:mm-N0p2baZSzxgfNbBcSi5HYBF67jdM7VZlJfTbhI2ht2cv1oCHRSL_FRPmM7DHr6ISpV91szCTOEg)
+   2. [Differentiating Communication Styles of Leaders on the Linux Kernel Mailing List](https://dl.acm.org/doi/abs/10.1145/2957792.2957801?casa_token=VMchS_jhea0AAAAA:EubJDL_ftM5jmV3_yzwWzDLvLq8hAsexZnss1x3j754OZr4VNENST_tSl0ijQEBnVg5AaFWpZGf3kQ)
+   3. [An Empirical Study on the Challenges of eBPF Application Development](https://dl.acm.org/doi/abs/10.1145/3672197.3673429)
\ No newline at end of file
diff --git a/Agent/README.md b/agent/README.txt
similarity index 100%
rename from Agent/README.md
rename to agent/README.txt