feat(controller): expose AgentTeams metrics

[shiyiyue1102]

Summary

• expose controller-runtime metrics with low-cardinality agentteams_* controller metrics
• add CR count collection and HTTP API route/status metrics
• add Helm metrics port plus optional ServiceMonitor, configured by AGENTTEAMS_METRICS_BIND_ADDR

Scope

• Controller metrics and Helm observability wiring only.
• Does not include AppService identity changes, Sandbox runtime changes, QwenPaw runtime, TeamHarness plugin runtime, or broad controller/Helm environment rename.

Validation

• go test ./internal/metrics ./internal/server ./internal/config ./internal/controller ./internal/app from hiclaw-controller
• git diff --cached --check

Local helm lint was not run because helm is not installed in this environment.

[github-actions]

📊 CI Metrics Report

Summary

Metric	Current	Baseline	Change
LLM Calls	157	81	+76 ↑ +93.8%
Input Tokens	3988019	2803871	+1184148 ↑ +42.2%
Output Tokens	37289	16791	+20498 ↑ +122.1%
Total Tokens	4025308	2820662	+1204646 ↑ +42.7%

By Role

Role	Metric	Current	Baseline	Change
🧠 Manager	LLM Calls	90	68	+22 ↑ +32.4%
	Input Tokens	2353484	2502214	-148730 ↓ -5.9%
	Output Tokens	17682	13725	+3957 ↑ +28.8%
	Total Tokens	2371166	2515939	-144773 ↓ -5.8%
🔧 Workers	LLM Calls	67	13	+54 ↑ +415.4%
	Input Tokens	1634535	301657	+1332878 ↑ +441.9%
	Output Tokens	19607	3066	+16541 ↑ +539.5%
	Total Tokens	1654142	304723	+1349419 ↑ +442.8%

Per-Test Breakdown

Test	Mgr Calls	Wkr Calls	Δ Calls	Mgr In	Wkr In	Mgr Out	Wkr Out	Δ Tokens	Trend
02-create-worker	4	0	-8 ↓ -66.7%	103977	0	458	0	-254187 ↓ -70.9%	✅ improved
03-assign-task	16	8	+9 ↑ +60.0%	427983	180786	2168	1786	+139067 ↑ +29.4%	⚠️ regressed
04-human-intervene	12	21	+20 ↑ +153.8%	279757	460123	1449	3053	+311384 ↑ +71.9%	⚠️ regressed
05-heartbeat	10	5	+8 ↑ +114.3%	234176	124781	3109	1199	+88013 ↑ +32.0%	⚠️ regressed
06-multi-worker	48	33	+47 ↑ +138.2%	1307591	868845	10498	13569	+920369 ↑ +71.9%	⚠️ regressed

Trends

✅ 1 test(s) improved (fewer LLM calls)
⚠️ 4 test(s) regressed (more LLM calls)

---

Generated by HiClaw CI on 2026-07-04 11:58:36 UTC

---

📦 Download debug logs & test artifacts

[shiyiyue1102]

@maplefeng-a The checks are passing now; could you help review this when you have time?

[maplefeng-a]

Looks good to me. I reviewed the controller metrics implementation and the relevant tests; CI is green. Approving.