Document: explicitly use anvilmcp sandbox for higher-risk scenarios
This commit is contained in:
parent
7c7d2de1e3
commit
0e4848f72f
|
|
@ -178,6 +178,8 @@ Validation is basic pattern matching. Not comprehensive - relies on sandbox for
|
|||
|
||||
By default (when the `sandbox` argument is not specified), `execute_code` uses the restricted `anvilmcp` sandbox configuration. If this fails (e.g., due to kernel or permission constraints), it falls back to the `default` sandbox — the same configuration used by the outer landrun invocation. This compromise provides "just works" behavior and avoids surprises, while still applying the tighter sandbox when possible.
|
||||
|
||||
For higher-risk scenarios (e.g., running untrusted scripts, processing sensitive data), the `anvilmcp` sandbox should be explicitly specified to ensure the tighter restrictions are enforced — it uses a temporary workspace, minimal filesystem access, and a restricted set of environment variables.
|
||||
|
||||
### Adjusting Timeout
|
||||
|
||||
Default: 30 seconds
|
||||
|
|
|
|||
|
|
@ -318,6 +318,8 @@ import { evil } from "/etc/passwd";
|
|||
|
||||
By default (when the `sandbox` argument is not specified), `execute_code` uses the restricted `anvilmcp` sandbox configuration. If this fails (e.g., due to kernel or permission constraints), it falls back to the `default` sandbox — the same configuration used by the outer landrun invocation. This compromise provides "just works" behavior and avoids surprises, while still applying the tighter sandbox when possible.
|
||||
|
||||
For higher-risk scenarios (e.g., running untrusted scripts, processing sensitive data), the `anvilmcp` sandbox should be explicitly specified to ensure the tighter restrictions are enforced — it uses a temporary workspace, minimal filesystem access, and a restricted set of environment variables.
|
||||
|
||||
### bash Permissions
|
||||
|
||||
```bash
|
||||
|
|
|
|||
Reference in New Issue