Files
codebase-memory-mcp/scripts
Andrew Hundt 82009eefeb test(bench): measure invocation overhead
Add optional overhead probes to the incremental benchmark so CLI and persistent MCP runs can measure cheap existing tool-call latency before indexing. The default remains zero probes, preserving historical gate behavior unless explicitly requested.

Validation: python3 -m py_compile scripts/benchmark-incremental-speed.py; git diff --check; bash scripts/check-source-safety.sh; scripts/benchmark-incremental-speed.py --transport cli --overhead-probes 5 --include-logs; scripts/benchmark-incremental-speed.py --transport mcp --overhead-probes 5 --include-logs.
Signed-off-by: Andrew Hundt <ATHundt@gmail.com>
2026-07-02 02:12:31 -04:00
..
2026-06-09 11:57:03 +02:00