Update section Conclusion

docs/6_exploitation.md

@@ -23,17 +23,10 @@
         3. [Run Final Exploit](./6_exploitation.md#run-final-exploit)
     4. [Conclusion](./6_exploitation.md#conclusion)
 <!--TODO--------------------------------------------------------------------------------------------
-- [X] Can we integrate morion/circled.rop3.py to circled.server.py?
-- [X] Try out manual exploit
 - [ ] Do in all chapters: ``` -> ```shell or ```python
-- [ ] Table of Contents
-- [ ] Can morion_rop_generator merge the payloads?
-- [ ] Document gdb debug output to show payload works
 - [ ] Create screencast using morion_pwndbg
 - [ ] Remove circled.exploit.md and circled.exploit.py from git repository
-- [ ] Why do we have `var:s+0`?
 - [ ] Recheck links [Symbolic Execution: Analysis Modules](./4_symbex.md#analysis-modules)
-- [X] Update figure references (e.g. Figure 5.4)
 - [ ] Morion README: Intended usage - crash triage
 --------------------------------------------------------------------------------------------------->
 # Exploitation
@@ -805,14 +798,21 @@ If the final exploit worked, you will receive a reverse shell on the targeted de
 attaching of the _gdbserver_, i.e. without the flag `--gdb`. You will receive the reverse shell,
 once the correct stack address of the system command has been found.
 ## Conclusion
-- Note: The shown exploit could easily be generated without using symbolic execution. However, we
-    have chosen it since it is rather easy to follow along and suitable to explain how Morion works.
+This repository intended to demonstrate (some of) the current functionalities (and limitations) of
+the PoC tool [Morion](https://github.com/pdamian/morion) and to enable interested readers to
+replicate all listed steps, fostering independent experimentation with the tool. It aimed to show
+how symbolic execution might assist in understanding the characteristics of a bug, determine its
+capabilities to decide whether or not a bug is exploitable, and if so support in the process of
+crafting an exploit. We showcased the tool on a real-world vulnerability to proof its
+practicability, however want to emphasize that the shown vulnerability class, stack-based buffer
+overflows, are well studied and rather easy to exploit. This might not necessarily hold true for
+other types of vulnerability classes whose exploitation might sometimes be more of an art than
+science, and in consequence typically more difficult to address by an (automated) tool.
 
-- Conclusions
-  - Well-known and rather simple to exploit vulnerability class (stack buffer overflow)
-  - Exploitation of others might be harder to automate with symbolic execution (heap overflows, race conditions, etc.)
-  - A lot open challenges regarding environment modeling / (semantic) function modeling (see my presentations)
-    - Sometimes not needed, sometimes a simplified model might work, sometimes minor details mather
+[Morion](https://github.com/pdamian/morion), but also symbolic execution in more general terms,
+still have a lot of limitations (scalability, modeling interactions with the environment, etc.),
+especially when being applied to real-world binaries. [Morion](https://github.com/pdamian/morion)
+was created with the intention to research and better understand these limits.
 
 ----------------------------------------------------------------------------------------------------
 [Back-to-Top](./6_exploitation.md#table-of-contents)