docs(team-leader): simplify worker name guidance

[RerankerGuo]

Summary

• replace defensive assignedTo warning blocks with a concise Matrix-localpart rule
• point DAG and Loop guidance at the same rule without duplicating examples
• simplify task-management Worker Name Canonicality while preserving hiclaw get workers / matrixUserID lookup guidance

Fixes #913.

Tests

• git diff --check
• rg -n "❌|mandatory before|Common mistake|Do NOT use CLI|strip prefixes|matrixUserID|assignedTo|assigned_to" manager/agent/team-leader-agent/skills/project-management/SKILL.md manager/agent/team-leader-agent/skills/project-management/references/dag-execution.md manager/agent/team-leader-agent/skills/task-management/SKILL.md

[github-actions]

📊 CI Metrics Report

Summary

Metric	Current	Baseline	Change
LLM Calls	116	81	+35 ↑ +43.2%
Input Tokens	3132781	2803871	+328910 ↑ +11.7%
Output Tokens	28957	16791	+12166 ↑ +72.5%
Total Tokens	3161738	2820662	+341076 ↑ +12.1%

By Role

Role	Metric	Current	Baseline	Change
🧠 Manager	LLM Calls	73	68	+5 ↑ +7.4%
	Input Tokens	2281648	2502214	-220566 ↓ -8.8%
	Output Tokens	15547	13725	+1822 ↑ +13.3%
	Total Tokens	2297195	2515939	-218744 ↓ -8.7%
🔧 Workers	LLM Calls	43	13	+30 ↑ +230.8%
	Input Tokens	851133	301657	+549476 ↑ +182.2%
	Output Tokens	13410	3066	+10344 ↑ +337.4%
	Total Tokens	864543	304723	+559820 ↑ +183.7%

Per-Test Breakdown

Test	Mgr Calls	Wkr Calls	Δ Calls	Mgr In	Wkr In	Mgr Out	Wkr Out	Δ Tokens	Trend
02-create-worker	5	0	-7 ↓ -58.3%	135200	0	827	0	-222595 ↓ -62.1%	✅ improved
03-assign-task	12	4	+1 ↑ +6.7%	332258	89255	1957	774	-49412 ↓ -10.4%	⚠️ regressed
04-human-intervene	19	22	+28 ↑ +215.4%	474338	335842	3296	2049	+382527 ↑ +88.3%	⚠️ regressed
05-heartbeat	8	2	+3 ↑ +42.9%	249966	43770	1494	270	+20248 ↑ +7.4%	⚠️ regressed
06-multi-worker	29	15	+10 ↑ +29.4%	1089886	382266	7973	10317	+210308 ↑ +16.4%	⚠️ regressed

Trends

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

---

Generated by HiClaw CI on 2026-07-02 10:59:42 UTC

---

📦 Download debug logs & test artifacts

[github-actions]

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

Commit: c9a0fad
Workflow run: #1404

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: 6ec6ee6
Workflow run: #1505

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: 6ec6ee6
Workflow run: #1505

Test Results

No test output captured.

Debug Log (tail)

No debug logs available.

---

📦 Download full debug logs & test artifacts

[RerankerGuo]

@shiyiyue1102 Hi! Both SHARD_A copaw/copaw and openclaw/openclaw failed. This is a docs-only PR (+14/-64 lines), so it's impossible for this change to affect runtime behavior — and the failures hit three different runtime combinations across multiple PRs (copaw/hermes on #984 / #972 / #973, copaw/copaw on #984, openclaw/openclaw on this PR), confirming a test infrastructure flake rather than any branch-local regression.

The SHARD_A test that's failing is test-06-multi-worker; on the #984 run I downloaded, Manager never invoked hiclaw create worker --name bob (no hiclaw-worker-bob session directory at all) — Manager only created alice. Could you re-run CI or merge this docs-only PR if you agree the flake is unrelated?