7 lines
246 B
Plaintext
7 lines
246 B
Plaintext
# Terminal-Bench regression task subset for Letta Code
|
|
# These tasks are run on a schedule to detect regressions.
|
|
# Criteria: fast (<10 min), diverse capabilities, deterministic.
|
|
# Adjust based on known Letta Code pass rates.
|
|
|
|
cancel-async-tasks
|