fix(config): pass custom model parameters to controller

[RerankerGuo]

Summary

• pass custom model context/max-token/reasoning/vision env vars into the embedded controller container
• expose the same model parameter overrides in Helm values and render them into the runtime Secret
• teach controller config to read reasoning/vision overrides and forward all model parameters to agentconfig

Fixes #961.

Tests

• bash -n install/hiclaw-install.sh

• git diff --check

• Go 1.25.0: go test ./internal/config
  Not run: helm template ... because Helm is not installed (helm: command not found).

Temporary dependency

• This branch is temporarily stacked on test: make multi-worker bob setup deterministic #985 (78e9385), whose full CI passed all 10 integration matrices.
• Relative to test: make multi-worker bob setup deterministic #985, this PR retains only its original scoped changes; the shared test fix disappears from the diff after test: make multi-worker bob setup deterministic #985 merges.
• Fresh checks on 4fc9eaa4 passed every build and all 10 integration matrices.

[github-actions]

📊 CI Metrics Report

Summary

Metric	Current	Baseline	Change
LLM Calls	130	81	+49 ↑ +60.5%
Input Tokens	3806369	2803871	+1002498 ↑ +35.8%
Output Tokens	35958	16791	+19167 ↑ +114.2%
Total Tokens	3842327	2820662	+1021665 ↑ +36.2%

By Role

Role	Metric	Current	Baseline	Change
🧠 Manager	LLM Calls	97	68	+29 ↑ +42.6%
	Input Tokens	3044708	2502214	+542494 ↑ +21.7%
	Output Tokens	20667	13725	+6942 ↑ +50.6%
	Total Tokens	3065375	2515939	+549436 ↑ +21.8%
🔧 Workers	LLM Calls	33	13	+20 ↑ +153.8%
	Input Tokens	761661	301657	+460004 ↑ +152.5%
	Output Tokens	15291	3066	+12225 ↑ +398.7%
	Total Tokens	776952	304723	+472229 ↑ +155.0%

Per-Test Breakdown

Test	Mgr Calls	Wkr Calls	Δ Calls	Mgr In	Wkr In	Mgr Out	Wkr Out	Δ Tokens	Trend
02-create-worker	6	0	-6 ↓ -50.0%	160202	0	899	0	-197521 ↓ -55.1%	✅ improved
03-assign-task	13	6	+4 ↑ +26.7%	371510	142471	2636	831	+43792 ↑ +9.2%	⚠️ regressed
04-human-intervene	23	12	+22 ↑ +169.2%	485784	231863	3212	1491	+289352 ↑ +66.8%	⚠️ regressed
05-heartbeat	11	1	+5 ↑ +71.4%	349875	25435	1920	90	+102068 ↑ +37.1%	⚠️ regressed
06-multi-worker	44	14	+24 ↑ +70.6%	1677337	361892	12000	12879	+783974 ↑ +61.2%	⚠️ regressed

Trends

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

---

Generated by HiClaw CI on 2026-07-13 15:35:47 UTC

---

📦 Download debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction / mgr=copaw / wk=hermes)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[RerankerGuo]

@shiyiyue1102 Hi! SHARD_A copaw/hermes failed on this PR's CI. Same pattern as #984 — in the copaw/hermes case the hiclaw-worker-bob session directory does not exist at all and Manager never invoked hiclaw create worker --name bob — Manager only created alice. Failure mode is a test-prompt interpretation flake, not the model-parameter env wiring in this PR.

This PR only changes Helm values + embedded install scripts + controller config to pass custom model context window / max tokens / reasoning / vision env vars. It doesn't touch the Manager agent runtime or worker-create code path. Could you re-run CI or merge if you agree the flake is unrelated?

[github-actions]

❌ Integration Tests Failed (controller-cr / mgr=openclaw / wk=openclaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (controller-cr / mgr=copaw / wk=hermes)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction-2 / mgr=openclaw / wk=openclaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (controller-cr / mgr=copaw / wk=copaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction-2 / mgr=copaw / wk=hermes)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction-2 / mgr=copaw / wk=copaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (controller-cr-2 / mgr=copaw / wk=copaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction / mgr=openclaw / wk=openclaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[github-actions]

❌ Integration Tests Failed (llm-interaction / mgr=copaw / wk=copaw)

Commit: 38c4993
Workflow run: #1650

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[shiyiyue1102]

Passing model context, token, reasoning, and vision overrides through every deployment path has product value, but the current PR is broad and conflicts with main. It also overlaps newer model-vision and QwenPaw work. Please rebase and verify which pieces are still missing, then either narrow this PR to one coherent model-parameter contract or split installer and controller changes if they can land independently. The updated branch should cover Helm, embedded Bash, PowerShell, config parsing, and nil versus explicit-false behavior with focused tests.