CalvinBackup/gstack
1
0
Fork 0
mirror of https://github.com/garrytan/gstack.git synced 2026-07-18 21:47:20 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix(test-harness): measure first response on first NDJSON line

Browse Source
This commit is contained in:
t
2026-07-14 12:56:07 -07:00
parent 7c9df1c568
commit 5cb8761d36
1 changed files with 6 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
test/helpers/session-runner.ts
+6 -2
View File
@@ -233,6 +233,11 @@ export async function runSkillTest(options: {
if (!line.trim()) continue; if (!line.trim()) continue;
collectedLines.push(line); collectedLines.push(line);
// Track time to first NDJSON line (measures latency from spawn to first Claude response)
if (firstResponseMs === 0) {
firstResponseMs = Date.now() - startTime;
}
// Real-time progress to stderr + persistent logs // Real-time progress to stderr + persistent logs
try { try {
const event = JSON.parse(line); const event = JSON.parse(line);
@@ -244,8 +249,7 @@ export async function runSkillTest(options: {
liveToolCount++; liveToolCount++;
const now = Date.now(); const now = Date.now();
const elapsed = Math.round((now - startTime) / 1000); const elapsed = Math.round((now - startTime) / 1000);
// Track timing telemetry // Track inter-turn latency (tool call to tool call)
if (firstResponseMs === 0) firstResponseMs = now - startTime;
if (lastToolTime > 0) { if (lastToolTime > 0) {
const interTurn = now - lastToolTime; const interTurn = now - lastToolTime;
if (interTurn > maxInterTurnMs) maxInterTurnMs = interTurn; if (interTurn > maxInterTurnMs) maxInterTurnMs = interTurn;