Agents load testing results

.scripts/sync-skills.py

@@ -113,6 +113,18 @@ def sync_template(template: str, config: dict):
             copy_skill(SOURCE / "agent-openai-memory", dest / "agent-memory")
 
 
+def sync_load_testing():
+    """Sync the load-testing skill to agent-load-testing (standalone, not a standard template)."""
+    template = "agent-load-testing"
+    template_path = REPO_ROOT / template
+    if not template_path.exists():
+        print(f"Skipping {template} (directory not found)")
+        return
+    print(f"Syncing {template}...")
+    dest = template_path / ".claude" / "skills" / "load-testing"
+    copy_skill(SOURCE / "load-testing", dest)
+
+
 def main():
     """Sync skills to all templates."""
     for template, config in TEMPLATES.items():
@@ -122,6 +134,10 @@ def main():
             continue
         print(f"Syncing {template}...")
         sync_template(template, config)
+
+    # Sync standalone templates that aren't in TEMPLATES registry
+    sync_load_testing()
+
     print("Done!")
 
 

agent-load-testing/.claude/CLAUDE.md

@@ -0,0 +1,13 @@
+# Agent Load Testing
+
+This directory contains tools for load testing Databricks Apps to find their maximum QPS (queries per second).
+
+## Directory Structure
+
+- `mock-agent-app/` — Mock agent app (based on `agent-openai-agents-sdk`) with `MockAsyncOpenAI` client
+- `load-test-scripts/` — Load testing scripts (`run_load_test.py`, `locustfile.py`, `dashboard_template.py`)
+- `load-test-runs/` — Test results (one subdirectory per run, each with dashboard + CSV data)
+
+## Skills
+
+Use `/load-testing` for an interactive walkthrough of the full load testing workflow.