oyadoti

4 ms·

1. 1.▲
   Oya – Keep tool outputs away from the LLM to cut tokens and stop injection(github.com)

   2 points by oyadoti 4mo ago|1 comments
2. 2.▲
   Prompting doesn't work Software does(oya.ai)

   2 points by oyadoti 7mo ago|0 comments
3. 3.▲
   SafeSkill – We scanned 10K AI skills for code exploits and prompt injection(safeskill.dev)

   1 points by oyadoti 8mo ago|0 comments