feat(runtime): consume AgentTeams worker env

[shiyiyue1102]

Summary

• make shared runtime bootstrap normalize AGENTTEAMS_* worker, storage, controller, auth, CMS, and Matrix env vars
• keep existing HICLAW_* deployment inputs as compatibility fallbacks while runtime scripts consume AGENTTEAMS_* as the primary contract
• update OpenClaw Worker and CoPaw Worker storage/STS paths plus focused shell and CoPaw tests

Scope

This PR is intentionally limited to runtime env/storage consumption. It does not switch controller or Helm env injection, and it does not include QwenPaw, TeamHarness, AppService, Sandbox pool follow-ups, or Edge/local worker credential work.

Validation

• bash shared/tests/test-hiclaw-env.sh
• bash shared/tests/test-oss-credentials.sh
• bash -n shared/lib/hiclaw-env.sh shared/lib/oss-credentials.sh worker/scripts/worker-entrypoint.sh copaw/scripts/copaw-worker-entrypoint.sh shared/tests/test-hiclaw-env.sh shared/tests/test-oss-credentials.sh
• python3 -m py_compile copaw/src/copaw_worker/hooks/tools/taskflow.py
• git diff --cached --check

Note: local CoPaw pytest via uv is still blocked by python-olm==3.2.16 native libolm build on this macOS toolchain (missing cmake/gmake path then clang const assignment failure), so full pytest is left to Linux CI.

[github-actions]

📊 CI Metrics Report

Summary

Metric	Current	Baseline	Change
LLM Calls	110	81	+29 ↑ +35.8%
Input Tokens	2742873	2803871	-60998 ↓ -2.2%
Output Tokens	29661	16791	+12870 ↑ +76.6%
Total Tokens	2772534	2820662	-48128 ↓ -1.7%

By Role

Role	Metric	Current	Baseline	Change
🧠 Manager	LLM Calls	63	68	-5 ↓ -7.4%
	Input Tokens	1624675	2502214	-877539 ↓ -35.1%
	Output Tokens	11835	13725	-1890 ↓ -13.8%
	Total Tokens	1636510	2515939	-879429 ↓ -35.0%
🔧 Workers	LLM Calls	47	13	+34 ↑ +261.5%
	Input Tokens	1118198	301657	+816541 ↑ +270.7%
	Output Tokens	17826	3066	+14760 ↑ +481.4%
	Total Tokens	1136024	304723	+831301 ↑ +272.8%

Per-Test Breakdown

Test	Mgr Calls	Wkr Calls	Δ Calls	Mgr In	Wkr In	Mgr Out	Wkr Out	Δ Tokens	Trend
02-create-worker	3	0	-9 ↓ -75.0%	81620	0	490	0	-276512 ↓ -77.1%	✅ improved
03-assign-task	6	4	-5 ↓ -33.3%	141292	92016	949	595	-238804 ↓ -50.4%	✅ improved
04-human-intervene	17	15	+19 ↑ +146.2%	342626	271000	3132	1744	+185504 ↑ +42.8%	⚠️ regressed
05-heartbeat	4	4	+1 ↑ +14.3%	93770	111658	1029	535	-68260 ↓ -24.8%	⚠️ regressed
06-multi-worker	33	24	+23 ↑ +67.6%	965367	643524	6235	14952	+349944 ↑ +27.3%	⚠️ regressed

Trends

✅ 2 test(s) improved (fewer LLM calls)
⚠️ 3 test(s) regressed (more LLM calls)

---

Generated by HiClaw CI on 2026-07-04 11:50:48 UTC

---

📦 Download debug logs & test artifacts

[shiyiyue1102]

@maplefeng-a CI is green now. Could you take a look when convenient?

[maplefeng-a]

CI is green and the CoPaw-scoped change looks good under the agreed review policy. Approving.