Multi-Mission Tool Bench

Fits assistants whose users pile on related or changing requests in one session. The dynamic decision-tree grading that accepts several valid paths and scores path optimality is the part to borrow.

What it asks the agent to do

• handle a second, related request that arrives after the first
• switch to a new mission mid-dialogue and drop the old context
• reuse information from an earlier mission in a later one
• find the shortest valid tool path for each mission

What to borrow from it

case templatetaxonomygrading design

Caveats

Data is generated by a multi-agent pipeline with no real users, the repo has no license file, and the user side is scripted.

Details

Links

• Paper
• Github