Run a staged testing workflow that checks type errors first, executes unit and integration tests in parallel, and runs end-to-end tests only after they pass. Receive pass and fail counts with details about failures.
中文简介正在准备中。当前显示英语简介。
Comprehensive testing workflow - unit tests ∥ integration tests → E2E tests
安装前可以查看交给 AI 代理的指令内容。
Run comprehensive test suite with parallel execution.
┌─────────────┐ ┌───────────┐
│ diagnostics │ ──▶ │ arbiter │ ─┐
│ (type check)│ │ (unit) │ │
└─────────────┘ └───────────┘ │
├──▶ ┌─────────┐
┌───────────┐ │ │ atlas │
│ arbiter │ ─┘ │ (e2e) │
│ (integ) │ └─────────┘
└───────────┘
Pre-flight Parallel Sequential
(~1 second) fast tests slow tests
| # | Agent | Role | Execution |
|---|---|---|---|
| 1 | arbiter | Unit tests, type checks, linting | Parallel |
| 1 | arbiter | Integration tests | Parallel |
| 2 | atlas | E2E/acceptance tests | After 1 passes |
Before running tests, check for type errors - they often cause test failures:
tldr diagnostics . --project --format text 2>/dev/null | grep "^E " | head -10
Why diagnostics first?
If errors found: Fix them BEFORE running tests. Type errors usually mean tests will fail anyway.
If clean: Proceed to Phase 1.
For large test suites, find only affected tests:
tldr change-impact --session
# or for explicit files:
tldr change-impact src/changed_file.py
This returns which tests to run based on what changed. Skip this for small projects or when you want full coverage.
# Run both in parallel
Task(
subagent_type="arbiter",
prompt="""
Run unit tests for: [SCOPE]
Include:
- Unit tests
- Type checking
- Linting
Report: Pass/fail count, failures detail
""",
run_in_background=true
)
Task(
subagent_type="arbiter",
prompt="""
Run integration tests for: [SCOPE]
Include:
- Integration tests
- API tests
- Database tests
Report: Pass/fail count, failures detail
""",
run_in_background=true
)
# Wait for both
[Check TaskOutput for both]
Task(
subagent_type="atlas",
prompt="""
Run E2E tests for: [SCOPE]
Include:
- End-to-end flows
- Acceptance tests
- UI tests if applicable
Report: Pass/fail count, screenshots on failure
"""
)
User: /test
→ All unit + integration + E2E tests
User: /test authentication
→ Only auth-related tests
User: /test --quick
→ Only unit tests (skip integration and E2E)
User: /test the new payment feature
Claude: Starting /test workflow for payment feature...
Phase 0: Pre-flight diagnostics...
$ tldr diagnostics . --project --format text | grep "^E "
(no type errors found)
Phase 1: Running parallel tests...
┌─────────────────────────────────────────┐
│ arbiter: Running unit tests... │
│ arbiter: Running integration tests... │
└─────────────────────────────────────────┘
arbiter: ✅ 45/45 unit tests passing
arbiter: ✅ 12/12 integration tests passing
Phase 2: Running E2E tests...
atlas: ✅ 8/8 E2E tests passing
Test Summary:
┌─────────────┬─────────┬────────┐
│ Type │ Passed │ Failed │
├─────────────┼─────────┼────────┤
│ Unit │ 45 │ 0 │
│ Integration │ 12 │ 0 │
│ E2E │ 8 │ 0 │
├─────────────┼─────────┼────────┤
│ TOTAL │ 65 │ 0 │
└─────────────┴─────────┴────────┘
All tests passing! ✅
User: /test
Claude: Starting /test workflow...
Phase 0: Pre-flight diagnostics...
$ tldr diagnostics . --project --format text | grep "^E "
E src/payment.py:45:12: Argument of type 'str' not assignable to 'int'
E src/refund.py:23:8: Return type 'None' not assignable to 'float'
Found 2 type errors. Fixing before running tests...
[Claude fixes the type errors]
Re-running diagnostics... clean.
Phase 1: Running parallel tests...
If Phase 1 fails:
arbiter: ❌ 43/45 tests passing
2 failures:
- test_payment_validation: expected 'invalid' got 'valid'
- test_refund_calculation: off by $0.01
Stopping workflow. Fix failures before running E2E tests.
--quick: Unit tests only--no-e2e: Skip E2E tests--coverage: Include coverage report--watch: Re-run on file changes还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Coordinate background agents through saved files, keeping lengthy work transcripts out of the main conversation. It defines how to pass results between agents, monitor progress, and verify completed work with tests.
中文简介正在准备中。当前显示英语简介。
Delegate implementation, file reading, and testing to separate agents, then bring concise summaries back to the main conversation. The rules explain when delegation helps and when a small edit is better handled directly.
中文简介正在准备中。当前显示英语简介。
Organize implementation work across research, planning, validation, implementation, and review agents. Each stage passes files to the next, while implementation starts with failing tests and the main conversation coordinates the workflow.
中文简介正在准备中。当前显示英语简介。
Connect Agentica SDK agents to Claude Code CLI through a proxy, an intermediary server. Use setup and debugging guidance to address file permissions, streamed responses, result formatting, and agents reporting success without performing the requested operations.
中文简介正在准备中。当前显示英语简介。
Consult Agentica APIs when building workflows where multiple AI agents cooperate. This reference covers eleven coordination patterns, agent tracking, shared state, handoffs, and ways to combine results, with pointers to detailed API specifications.
中文简介正在准备中。当前显示英语简介。
Write clearer instructions for Agentica and REPL agents by specifying roles, file operations, and output formats. Includes templates for dividing research, coordinating specialists, reviewing solutions, and passing results between agents through files.
中文简介正在准备中。当前显示英语简介。