test(cli): comprehensive test suite for is_complex_task, display_task_plan, and _make_agent (+80 tests)

[manus-use]

Summary

Comprehensive test suite for the three core CLI dispatch helper functions that have zero dedicated test coverage:

• is_complex_task — the heuristic that decides single-agent vs multi-agent routing
• display_task_plan — renders the multi-agent execution plan table via Rich
• _make_agent — factory that instantiates the correct agent class (browser/data/mcp/manus)

What's tested (80 tests)

is_complex_task (55 tests)

• Word count threshold (>30 words → complex)
• Sentence count threshold (>2 sentences split by . or ;)
• All 10 regex patterns: and...and, then, after, analyze...create/generate/build, compare...summarize, multiple, steps?, workflow, ordinals (first/second/third/finally), visualize/chart/graph...analyze/data, browse...extract/analyze, research...implement/create
• Case insensitivity verification
• Word boundary correctness (e.g. hereafter ≠ after, sandbox ≠ and, stepping ≠ steps)
• Edge cases: empty string, whitespace, special characters, multiline input, URLs with dots

display_task_plan (8 tests)

• Single task with no dependencies
• Multiple tasks with chained dependencies
• Long description truncation (>50 chars → ...)
• Exactly 50-char descriptions preserved
• Empty task list renders header
• Comma-separated dependency display
• All AgentType values (manus, browser, data_analysis, mcp)

_make_agent (17 tests)

• Correct agent class instantiation for each type (browser → BrowserUseAgent, data → DataAnalysisAgent, mcp → MCPAgent, manus/default → ManusAgent)
• Unknown/empty agent type falls back to ManusAgent
• Keyword argument forwarding (callback_handler, multiple kwargs)
• Exception propagation (ImportError, RuntimeError, TypeError)

Test quality

• 100% mocked — no real HTTP calls or agent instantiation
• All tests isolated and deterministic
• Follows project conventions (pytest, unittest.mock)
• Clean ruff check + ruff format

Test results

1238 passed, 3 deselected, 3 warnings in 24.46s

(Baseline 1158 + 80 new tests, 0 failures)

Checked against open PRs (no overlap)

• tests: comprehensive CLI tests for _run_analyze, _run_remediate, _run_discover (+78 tests) #123 (tests for _run_analyze, _run_remediate, _run_discover — different functions)
• test(cli): comprehensive test suite for manus-agent blast-radius subcommand #121 (tests for blast-radius CLI subcommand — different function)
• test(cli): comprehensive test suite for manus-agent history subcommand (+39 tests) #132 (tests for history CLI subcommand — different function)
• test(cli): comprehensive test suite for manus-agent poc-search subcommand (+57 tests) #139 (tests for poc-search CLI subcommand — different function)
• test(tools): comprehensive test suite for manus_workflow module (+58 tests) #150 (tests for manus_workflow module — different module)
• test(multi_agents): comprehensive test suite for orchestrator, workflow agent, and data types (+73 tests) #151 (tests for multi_agents module — different module)

No existing open or merged PR tests is_complex_task, display_task_plan, or _make_agent directly.