v1ktorrr0x/antigravity-filter-bypass ? reverse-engineered prompt

Reverse engineered prompt

Build me a small Python plugin for Antigravity that handles false stop events more gracefully.

I want it to watch for the Stop hook, check the last line of the latest model message, and if that line matches a known false refusal pattern, keep the agent moving instead of ending the task. If it is not a false trigger, it should leave normal behavior alone.

Please include the plugin manifest, the hook registration, the hook script, and a simple test file that covers the expected cases. Make sure the tests are easy to run locally with python test_filter_bypass.py and that everything passes. If you need to look up current docs online, go ahead.

Keep the behavior focused, only act on the final line of the last message, and make the recovery message clear enough that the agent knows to continue from the previous task without getting stuck.

Are you gonna build this?

make sure you review the code using coderabbit

Try freeSponsored — opens CodeRabbit in a new tab