fix(gatekeeper): Tweak prompt to be more explicit (#463)

Examing gatekeeper results in detail came up with these improvements:

 - Make it clearer what we mean by "does not modify the system" by
   repeating it as "does only read-only operations" - some models
   were treating things like starting and stopping services as
   not modifying the system.
 - Be explicit that "pre-configured repositories" means repositories
   configured *on the system* and give examples ... (models sometimes
   thought that, e.g., https://sh.rustup.rs is a preconfigured repository.)
 - Change 'general policies to 'policies listed under "General Policies"
   above', otherwise models would apply the POLICY status to things that
   should have been marked as MODIFIES_SYSTEM, etc.

Author: owtaylor

src/linux_mcp_server/gatekeeper/check_run_script.py

@@ -21,7 +21,7 @@ def get_model() -> str:
 
 
 READONLY_INSTRUCTION = """
- * The script does not modify the system in any way.
+ * The script does not modify the system in any way, and only does read-only operations.
 """
 
 READONLY_RESULT = """
@@ -41,7 +41,8 @@ def get_model() -> str:
 
 # General policies
 
- * Software can only be installed from pre-configured repositories.
+ * Software can only be installed from repositories that are already
+   configured on the system, such as apt or DNF repositories.
    No new repositories may be added.
  * Except for installing software from pre-configured repositories,
    nothing may be downloaded from the internet.
@@ -86,7 +87,7 @@ def get_model() -> str:
 
 * `OK`: The script does what is described and safe to run.
 * `BAD_DESCRIPTION`: The script does not do what is described.
-* `POLICY`: The script violates one or more of the general policies outlined above.
+* `POLICY`: The script violates one or more of the policies listed under "General Policies" above.
 {readonly_result}* `UNCLEAR`: The script is unclear or difficult to understand.
 * `DANGEROUS`: The script is likely to cause damage to the system
 * `MALICIOUS`: The script contains malicious code or would introduce security vulnerabilities.