Describe the bugaichat executes "may" types of commands without asking for permission
aichat --model openai:gpt-4o -r %functions% print the 123rd prime
Call may_execute_py_code {"code":"from sympy import primerange\nprime_list = list(primerange(0,730))\nprint(prime_list[122])"}
The 123rd prime number is 677.
I'm using the current main version of aichat (#64982b45) on WSL on Windows, along with commit a799428b397ac7789a91cf94a9c408ee8a2dd6e2 of llm-functions
Expected behavior
I expect to be asked whether to run the command.
argc version
Linux DESKTOP-DBPFNLP 5.15.146.1-microsoft-standard-WSL2 #1 SMP Thu Jan 11 04:09:03 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
argc 1.18.0
jq-1.6
/usr/bin/bash GNU bash, version 5.1.16(1)-release (x86_64-pc-linux-gnu)
/usr/bin/python Python 3.10.12
Describe the bug
aichat
executes "may" types of commands without asking for permissionI'm using the current main version of aichat (#64982b45) on WSL on Windows, along with commit a799428b397ac7789a91cf94a9c408ee8a2dd6e2 of llm-functions
Expected behavior I expect to be asked whether to run the command.
Screenshots/Logs
Environment