This is an automated email from the ASF dual-hosted git repository.
davsclaus pushed a change to branch main
in repository https://gitbox.apache.org/repos/asf/camel-performance-tests.git
from 927fdb0 ai-benchmark: the ladder keeps values in variables, not
exchange properties
new c46e40a ai-benchmark: BENCH_FAST waits for what a step checks instead
of fixed sleeps, and traces the tool times
new 4fae7f9 ai-benchmark: BENCH_FAST counts only the errors after a step
began, BENCH_SERVER_INSTRUCTIONS, and data dirs for the ladder
new d889c91 ai-benchmark: 25 tool calls per step instead of 20
new f262f3e ai-benchmark: the server restart stops only its own server,
and the postgres wait reads the new JSON errors
new 6036d63 ai-benchmark: checks accept quoted values, and the apps scan
every 500 ms like camel_run
new fe437d5 ai-benchmark: camel_catalog_doc answers are cut at 16000
characters, and the trace records each answer's length
The 6 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "add" were already present in the repository and have only
been added to this reference.
Summary of changes:
ai-benchmark/agent_local.py | 5 +-
ai-benchmark/agent_mcp_stepwise.py | 175 ++++++++++++++++++++++++++++++++-----
ai-benchmark/gen_stepwise.py | 15 ++--
ai-benchmark/start-server.sh | 7 +-
4 files changed, 170 insertions(+), 32 deletions(-)
