chore: add better docs for safe_mode; load semgrep if available #709
Conversation
| Install Open Interpreter with the safety toolkit dependencies as part of the bundle: | ||
|
|
||
| ```shell | ||
| pip install open-interpreter[safe] |
There was a problem hiding this comment.
We can leverage the tool.poetry.extras configuration to allow folks to optionally install semgrep and other safety tools via the [safe] extra.
There was a problem hiding this comment.
YES. I love this idea.
I think it would be great to make a slimmed down version of OI called [core] (wouldn't even include rich or any terminal interface), [safe] for semgrep, perhaps docker, etc, then eventually [all] if we build features beyond safe mode with dependencies.
| interpreter_intro_message.append(f"**Safe Mode**: {interpreter.safe_mode}\n\n>Note: **Safe Mode** requires `semgrep` (`pip install semgrep`)") | ||
| if interpreter.safe_mode == "ask" or interpreter.safe_mode == "auto": | ||
| if not check_for_package("semgrep"): | ||
| interpreter_intro_message.append(f"**Safe Mode**: {interpreter.safe_mode}\n\n>Note: **Safe Mode** requires `semgrep` (`pip install semgrep`)") |
There was a problem hiding this comment.
Now we only need to show the semgrep message if it isn't already in the user's current environment. If semgrep is available and safe_mode is enabled, semgrep will be automatically loaded.
|
Great work Eric. Really happy to start seeing [extras] in OI— opens the door to OI serving different kinds of users more precisely. |
…afe-mode-docs chore: add better docs for safe_mode; load semgrep if available Former-commit-id: d588a1d Former-commit-id: 52a0e45f954a7492dcbbf298ac1bc8a481edec44 Former-commit-id: 424868a7bcadab255ac122367bb31c8b1898acde [formerly d23999f692e636d257530028f7ddac48c537e4ca] Former-commit-id: 4013995fcb4278e520468ec2c60e0f6fad008ca2
…afe-mode-docs chore: add better docs for safe_mode; load semgrep if available
Describe the changes you have made:
Update docs for
safe_mode, dynamically loadsemgrepif it's available.Reference any relevant issue (Fixes #000)
I have tested the code on the following OS:
AI Language Model (if applicable)