Tool Use updates

Author: elifarley

README.md

@@ -232,17 +232,23 @@ There are [many more examples](test/corpus) to look at...
 
 ### Use as a refactoring language / _diff_ format
 
-One can use `CEDARScript` to concisely and unambiguously represent code modifications at a higher level than a standard diff format can.
+One can use `CEDARScript` to _concisely_ and _unambiguously_ represent code modifications at a _higher_ level than a standard `diff` format can.
 
-IDEs can store the local history of files in CEDARScript format, and this can also be used for searches.
+IDEs can store the local history of files in `CEDARScript` format, and this can also be used for searches.
 
 ### Tool Use
-If **explicit** configuration is set, the [**CEDARScript runtime**](https://github.com/CEDARScript/cedarscript-editor-python) can act as
-a gateway through which an LLM can:
-1. Call local commands (`ls`, `grep`, `find`, `open`)
-2. Run scripts
-3. Call external HTTP API services
-4. See the user's screen and take control of the mouse and keyboard
+If **explicit** configuration is set, the [**CEDARScript runtime**](https://github.com/CEDARScript/cedarscript-editor-python)
+can act as a **unified gateway** through which _any_ LLM can call external commands and obtain their output
+(a.k.a. **Tool Use** support).
+
+This includes:
+1. **Web browsing**
+2. **Code Interpreter**
+   - Run scripts written in Python, Bash, Javascript, Lua, etc
+3. **Function Calling**
+    - Call local commands (`ls`, `grep`, `find`, `open`)
+    - Call external HTTP API services
+4. **Computer Use**: See the user's screen and take control of the mouse and keyboard
 5. Possibilities are numerous...
 
 The output from the external tool is captured and sent back to the LLM.

proposals/capture-external-command-output/README.md

@@ -1,13 +1,23 @@
 # External command calls to allow LLMs to experiment and learn
 
 ## Summary
-I propose a language extension to `CEDARScript` to allow arbitrary command execution.
+I propose a language extension to `CEDARScript` to allow arbitrary command execution, a.k.a. **Tool Use** support.
 It would work as an escape hatch that connects the LLM to the external environment.
 
+This includes:
+1. **Web browsing**
+2. **Code Interpreter**
+    - Run scripts written in Python, Bash, Javascript, Lua, etc
+3. **Function Calling**
+    - Call local commands (`ls`, `grep`, `find`, `open`)
+    - Call external HTTP API services
+4. **Computer Use**: See the user's screen and take control of the mouse and keyboard
+5. Possibilities are numerous...
+
 Security implications aside, it holds a lot of potential.
 
-The firs application could be a benchmark to see if the LLM would be able to perform better than the baseline
-when solving complex mathematical challenges, e. g. https://epoch.ai/frontiermath
+The first application could be a benchmark to see if the LLM would be able to perform better than the baseline
+when solving complex mathematical challenges, e.g. https://epoch.ai/frontiermath
 
 ## Examples