Add synthetic test data, MCP registry manifest, risk-based test selection

README.md

@@ -13,6 +13,7 @@
 
 ## Table of Contents
 
+- [What's new (2026-06-19) — Test & Tooling Batch](#whats-new-2026-06-19--test--tooling-batch)
 - [What's new (2026-06-19) — Transactional Queue](#whats-new-2026-06-19--transactional-queue)
 - [What's new (2026-06-19) — Unattended Reliability](#whats-new-2026-06-19--unattended-reliability)
 - [What's new (2026-06-19) — Popup Watchdog](#whats-new-2026-06-19--popup-watchdog)
@@ -63,6 +64,14 @@
 
 ---
 
+## What's new (2026-06-19) — Test & Tooling Batch
+
+Three pure-stdlib quality-of-life tools, full stack (facade, `AC_*`, MCP, Script Builder). Full reference: [`docs/source/Eng/doc/new_features/v11_features_doc.rst`](docs/source/Eng/doc/new_features/v11_features_doc.rst).
+
+- **Synthetic test data** — `generate_rows(schema, count, seed=...)` / `write_dataset` (`AC_generate_data`, `ac_generate_data`): deterministic fake rows (name/email/phone/int/choice/date…) to drive data-driven runs without real PII; no Faker.
+- **MCP registry manifest** — `write_server_manifest("server.json", include_tools=True)` (`AC_mcp_manifest`, `ac_mcp_manifest`): publish a registry-valid `server.json` so MCP agents/IDEs can discover this server.
+- **Risk-based test selection** — `rank_flows` / `select_flows` (`AC_rank_tests` / `AC_select_tests`): rank flows by recent failures, flakiness, staleness and never-run from run history; run the riskiest first or only the top-k.
+
 ## What's new (2026-06-19) — Transactional Queue
 
 Turn AutoControl from "run a script" into "run a robot." A SQLite-backed work queue implements the production-RPA dispatcher/performer pattern: enqueue items, process one at a time with per-item status, dedup and retry, so a run of thousands is **resumable after a crash** and parallelizable. Pure stdlib, full stack. Full reference: [`docs/source/Eng/doc/new_features/v10_features_doc.rst`](docs/source/Eng/doc/new_features/v10_features_doc.rst).

README/README_zh-CN.md

@@ -12,6 +12,7 @@
 
 ## 目录
 
+- [本次更新 (2026-06-19) — 测试与工具三件套](#本次更新-2026-06-19--测试与工具三件套)
 - [本次更新 (2026-06-19) — 事务式工作队列](#本次更新-2026-06-19--事务式工作队列)
 - [本次更新 (2026-06-19) — 无人值守可靠性](#本次更新-2026-06-19--无人值守可靠性)
 - [本次更新 (2026-06-19) — 弹窗看门狗](#本次更新-2026-06-19--弹窗看门狗)
@@ -62,6 +63,14 @@
 
 ---
 
+## 本次更新 (2026-06-19) — 测试与工具三件套
+
+三项纯标准库的生产力工具,走完整五层(facade、`AC_*`、MCP、Script Builder)。完整参考:[`docs/source/Zh/doc/new_features/v11_features_doc.rst`](../docs/source/Zh/doc/new_features/v11_features_doc.rst)。
+
+- **合成测试数据** — `generate_rows(schema, count, seed=...)` / `write_dataset`(`AC_generate_data`、`ac_generate_data`):生成可重现的假数据行(name/email/phone/int/choice/date…),驱动数据驱动执行而不需真实 PII;无需 Faker。
+- **MCP registry 清单** — `write_server_manifest("server.json", include_tools=True)`(`AC_mcp_manifest`、`ac_mcp_manifest`):生成符合 registry 规范的 `server.json`,让 MCP agent/IDE 能发现此服务器。
+- **风险导向测试选择** — `rank_flows` / `select_flows`(`AC_rank_tests` / `AC_select_tests`):依最近失败、不稳定、陈旧与从未跑过,从 run history 排序流程;先跑最高风险或只跑前 k 个。
+
 ## 本次更新 (2026-06-19) — 事务式工作队列
 
 把 AutoControl 从「跑脚本」升级成「跑机器人」。以 SQLite 为底的工作队列实作生产级 RPA dispatcher/performer:入列项目、一次处理一项、具每项状态/去重/重试,使上千项执行能**崩溃后续跑**且可并行化。纯标准库、走完整五层。完整参考:[`docs/source/Eng/doc/new_features/v10_features_doc.rst`](../docs/source/Eng/doc/new_features/v10_features_doc.rst)。

README/README_zh-TW.md

@@ -12,6 +12,7 @@
 
 ## 目錄
 
+- [本次更新 (2026-06-19) — 測試與工具三件套](#本次更新-2026-06-19--測試與工具三件套)
 - [本次更新 (2026-06-19) — 交易式工作佇列](#本次更新-2026-06-19--交易式工作佇列)
 - [本次更新 (2026-06-19) — 無人值守可靠性](#本次更新-2026-06-19--無人值守可靠性)
 - [本次更新 (2026-06-19) — 彈窗看門狗](#本次更新-2026-06-19--彈窗看門狗)
@@ -62,6 +63,14 @@
 
 ---
 
+## 本次更新 (2026-06-19) — 測試與工具三件套
+
+三項純標準庫的生產力工具,走完整五層(facade、`AC_*`、MCP、Script Builder)。完整參考:[`docs/source/Zh/doc/new_features/v11_features_doc.rst`](../docs/source/Zh/doc/new_features/v11_features_doc.rst)。
+
+- **合成測試資料** — `generate_rows(schema, count, seed=...)` / `write_dataset`(`AC_generate_data`、`ac_generate_data`):產生可重現的假資料列(name/email/phone/int/choice/date…),驅動資料驅動執行而不需真實 PII;不需 Faker。
+- **MCP registry 清單** — `write_server_manifest("server.json", include_tools=True)`(`AC_mcp_manifest`、`ac_mcp_manifest`):產生符合 registry 規範的 `server.json`,讓 MCP agent/IDE 能發現此伺服器。
+- **風險導向測試選擇** — `rank_flows` / `select_flows`(`AC_rank_tests` / `AC_select_tests`):依最近失敗、不穩定、陳舊與從未跑過,從 run history 排序流程;先跑最高風險或只跑前 k 個。
+
 ## 本次更新 (2026-06-19) — 交易式工作佇列
 
 把 AutoControl 從「跑腳本」升級成「跑機器人」。以 SQLite 為底的工作佇列實作生產級 RPA dispatcher/performer:入列項目、一次處理一項、具每項狀態/去重/重試,使上千項執行能**當機後續跑**且可平行化。純標準庫、走完整五層。完整參考:[`docs/source/Zh/doc/new_features/v10_features_doc.rst`](../docs/source/Zh/doc/new_features/v10_features_doc.rst)。

docs/source/Eng/doc/new_features/v11_features_doc.rst

@@ -0,0 +1,73 @@
+==================================================
+New Features (2026-06-19) — Test & Tooling Batch
+==================================================
+
+Three quality-of-life tools, all pure standard library and wired through
+the full stack (facade, ``AC_*`` executor commands, MCP tools, Script
+Builder): seeded synthetic test data, an MCP registry ``server.json``
+generator, and risk-based test selection.
+
+.. contents::
+   :local:
+   :depth: 2
+
+
+Synthetic test data
+===================
+
+Generate deterministic fake rows from a tiny field schema — to drive
+data-driven runs without shipping real PII. No Faker dependency; the same
+``seed`` always yields the same rows::
+
+    from je_auto_control import generate_rows, write_dataset
+
+    rows = generate_rows({
+        "name": "name",
+        "email": {"type": "email", "domain": "acme.test"},
+        "age": {"type": "int", "min": 18, "max": 65},
+        "status": {"type": "choice", "choices": ["new", "vip"]},
+    }, count=100, seed=7)
+
+    write_dataset(rows, "people.csv")   # or .json
+
+Supported field types: ``first_name``, ``last_name``, ``name``,
+``username``, ``email``, ``phone``, ``city``, ``company``, ``word``,
+``sentence``, ``uuid``, ``bool``, ``int`` (min/max), ``float``
+(min/max/ndigits), ``choice`` (choices), ``date`` (start/end).
+
+The ``AC_generate_data`` command writes a file (then feed it to
+``AC_load_data``) or returns the rows inline.
+
+
+MCP registry manifest
+====================
+
+Publish a ``server.json`` describing this AutoControl MCP server so
+MCP-aware agents and IDEs can discover and install it. The manifest is
+built from live package metadata, so it never drifts::
+
+    from je_auto_control import write_server_manifest
+
+    write_server_manifest("server.json", include_tools=True)
+
+``include_tools`` embeds the live tool list under ``_meta`` (without
+touching the registry-valid core fields). Also exposed as
+``AC_mcp_manifest`` and the ``ac_mcp_manifest`` MCP tool.
+
+
+Risk-based test selection
+========================
+
+Instead of always running the whole suite, rank flows by how *risky* they
+are — recently failing, flaky, stale, or never-run — using the run-history
+store, then run the riskiest first (or only the top-k)::
+
+    from je_auto_control import select_flows, rank_flows
+
+    ranked = rank_flows(["login", "checkout", "report"])
+    risky = select_flows(["login", "checkout", "report"], k=2)
+
+The score is ``0.5*failure_rate + 0.2*last_failed + 0.2*flakiness +
+0.1*staleness``; a never-run flow scores ``0.8`` (untested is risky).
+Exposed as ``AC_rank_tests`` / ``AC_select_tests`` and the
+``ac_rank_tests`` / ``ac_select_tests`` MCP tools.

docs/source/Eng/eng_index.rst

@@ -33,6 +33,7 @@ Comprehensive guides for all AutoControl features.
    doc/new_features/v8_features_doc
    doc/new_features/v9_features_doc
    doc/new_features/v10_features_doc
+   doc/new_features/v11_features_doc
    doc/ocr_backends/ocr_backends_doc
    doc/observability/observability_doc
    doc/operations_layer/operations_layer_doc

docs/source/Zh/doc/new_features/v11_features_doc.rst

@@ -0,0 +1,72 @@
+============================================
+新功能 (2026-06-19) — 測試與工具三件套
+============================================
+
+三項提升生產力的工具,皆為純標準庫,並走完整五層(facade、``AC_*``
+執行器指令、MCP 工具、Script Builder):有種子的合成測試資料、MCP
+registry ``server.json`` 產生器,以及風險導向的測試選擇。
+
+.. contents::
+   :local:
+   :depth: 2
+
+
+合成測試資料
+============
+
+依一個極小的欄位 schema 產生**可重現**的假資料列——用來驅動資料驅動的
+執行,而不必散布真實 PII。不需要 Faker;相同的 ``seed`` 永遠產生相同的
+資料列::
+
+    from je_auto_control import generate_rows, write_dataset
+
+    rows = generate_rows({
+        "name": "name",
+        "email": {"type": "email", "domain": "acme.test"},
+        "age": {"type": "int", "min": 18, "max": 65},
+        "status": {"type": "choice", "choices": ["new", "vip"]},
+    }, count=100, seed=7)
+
+    write_dataset(rows, "people.csv")   # 或 .json
+
+支援的欄位型別:``first_name``、``last_name``、``name``、``username``、
+``email``、``phone``、``city``、``company``、``word``、``sentence``、
+``uuid``、``bool``、``int``(min/max)、``float``(min/max/ndigits)、
+``choice``(choices)、``date``(start/end)。
+
+``AC_generate_data`` 指令會寫出檔案(再交給 ``AC_load_data``)或直接
+回傳資料列。
+
+
+MCP registry 清單
+=================
+
+產生描述此 AutoControl MCP 伺服器的 ``server.json``,讓支援 MCP 的
+agent 與 IDE 能發現並安裝它。清單由即時套件中繼資料建構,因此不會與
+實際能力脫節::
+
+    from je_auto_control import write_server_manifest
+
+    write_server_manifest("server.json", include_tools=True)
+
+``include_tools`` 會把即時工具清單嵌入 ``_meta``(不更動 registry 規範
+的核心欄位)。同時提供 ``AC_mcp_manifest`` 與 ``ac_mcp_manifest`` MCP
+工具。
+
+
+風險導向測試選擇
+================
+
+與其每次都跑整套測試,不如依據流程的**風險**排序——最近失敗、不穩定
+(flaky)、太久沒跑、或從未跑過——用 run-history 紀錄計算,然後先跑最
+高風險的(或只跑前 k 個)::
+
+    from je_auto_control import select_flows, rank_flows
+
+    ranked = rank_flows(["login", "checkout", "report"])
+    risky = select_flows(["login", "checkout", "report"], k=2)
+
+分數為 ``0.5*失敗率 + 0.2*上次失敗 + 0.2*不穩定度 + 0.1*陳舊度``;
+從未跑過的流程得 ``0.8``(未測試即高風險)。提供 ``AC_rank_tests`` /
+``AC_select_tests`` 以及 ``ac_rank_tests`` / ``ac_select_tests`` MCP
+工具。

docs/source/Zh/zh_index.rst

@@ -33,6 +33,7 @@ AutoControl 所有功能的完整使用指南。
    doc/new_features/v8_features_doc
    doc/new_features/v9_features_doc
    doc/new_features/v10_features_doc
+   doc/new_features/v11_features_doc
    doc/ocr_backends/ocr_backends_doc
    doc/observability/observability_doc
    doc/operations_layer/operations_layer_doc

je_auto_control/__init__.py

@@ -113,6 +113,14 @@
 from je_auto_control.utils.work_queue import (
     BusinessError, WorkItem, WorkQueue,
 )
+# Seeded synthetic test-data generation
+from je_auto_control.utils.test_data import generate_rows, write_dataset
+# Risk-based test selection from run history
+from je_auto_control.utils.test_select import rank_flows, select_flows
+# MCP registry server.json manifest
+from je_auto_control.utils.mcp_registry import (
+    build_server_manifest, write_server_manifest,
+)
 # Background popup/interrupt watchdog (unattended automation)
 from je_auto_control.utils.watchdog import (
     PopupWatchdog, WatchdogRule, default_popup_watchdog,
@@ -514,6 +522,9 @@ def start_autocontrol_gui(*args, **kwargs):
     "handle_file_dialog", "FileDialogDriver",
     "ensure_interactive_session", "is_session_locked",
     "WorkQueue", "WorkItem", "BusinessError",
+    "generate_rows", "write_dataset",
+    "rank_flows", "select_flows",
+    "build_server_manifest", "write_server_manifest",
     # MCP server
     "AuditLogger", "HttpMCPServer", "MCPContent", "MCPPrompt",
     "MCPPromptArgument", "MCPResource", "MCPServer", "MCPTool",

je_auto_control/gui/script_builder/command_schema.py

@@ -653,6 +653,50 @@ def _add_misc_specs(specs: List[CommandSpec]) -> None:
         description="Fail if the session is locked / non-interactive.",
     ))
     _add_work_queue_specs(specs)
+    _add_tooling_specs(specs)
+
+
+def _add_tooling_specs(specs: List[CommandSpec]) -> None:
+    specs.append(CommandSpec(
+        "AC_generate_data", "Data", "Generate Synthetic Data",
+        fields=(
+            FieldSpec("count", FieldType.INT, optional=True, default=10),
+            FieldSpec("path", FieldType.FILE_PATH, optional=True),
+            FieldSpec("fmt", FieldType.ENUM, choices=("json", "csv"),
+                      optional=True),
+            FieldSpec("seed", FieldType.INT, optional=True),
+        ),
+        description="Generate seeded fake rows from a 'schema' (JSON view); "
+                    "writes a file when 'path' is set.",
+    ))
+    specs.append(CommandSpec(
+        "AC_mcp_manifest", "Tools", "MCP Registry Manifest",
+        fields=(
+            FieldSpec("path", FieldType.FILE_PATH, optional=True,
+                      default="server.json"),
+            FieldSpec("include_tools", FieldType.BOOL, optional=True,
+                      default=False),
+        ),
+        description="Write an MCP registry server.json for this server.",
+    ))
+    specs.append(CommandSpec(
+        "AC_rank_tests", "Testing", "Rank Tests by Risk",
+        fields=(
+            FieldSpec("history_path", FieldType.FILE_PATH, optional=True),
+            FieldSpec("window", FieldType.INT, optional=True, default=10),
+        ),
+        description="Score 'flows' (JSON view) by risk from run history.",
+    ))
+    specs.append(CommandSpec(
+        "AC_select_tests", "Testing", "Select Risky Tests",
+        fields=(
+            FieldSpec("k", FieldType.INT, optional=True),
+            FieldSpec("threshold", FieldType.FLOAT, optional=True),
+            FieldSpec("history_path", FieldType.FILE_PATH, optional=True),
+            FieldSpec("window", FieldType.INT, optional=True, default=10),
+        ),
+        description="Pick riskiest 'flows' (JSON view): top-k or threshold.",
+    ))
 
 
 def _add_work_queue_specs(specs: List[CommandSpec]) -> None:

je_auto_control/utils/executor/action_executor.py

@@ -2346,6 +2346,48 @@ def _queue_stats(db: str, name: str = "default") -> Dict[str, int]:
     return _queue(db, name).stats()
 
 
+def _generate_data(schema: Dict[str, Any], count: int = 10,
+                   path: Optional[str] = None, fmt: Optional[str] = None,
+                   seed: Optional[int] = None) -> Dict[str, Any]:
+    """Adapter: generate synthetic rows; write to ``path`` when given."""
+    from je_auto_control.utils.test_data import generate_rows, write_dataset
+    rows = generate_rows(schema, int(count), seed=seed)
+    if path:
+        return {"path": write_dataset(rows, path, fmt), "count": len(rows)}
+    return {"rows": rows, "count": len(rows)}
+
+
+def _mcp_manifest(path: Optional[str] = None,
+                  include_tools: bool = False) -> Dict[str, Any]:
+    """Adapter: build (or write) the MCP registry server.json manifest."""
+    from je_auto_control.utils.mcp_registry import (
+        build_server_manifest, write_server_manifest)
+    if path:
+        return {"path": write_server_manifest(
+            path, include_tools=bool(include_tools))}
+    return {"manifest": build_server_manifest(
+        include_tools=bool(include_tools))}
+
+
+def _rank_tests(flows: List[str], history_path: Optional[str] = None,
+                window: int = 10) -> Dict[str, Any]:
+    """Adapter: score flows by risk (riskiest first)."""
+    from je_auto_control.utils.test_select import rank_flows
+    return {"ranked": rank_flows(flows, history_path=history_path,
+                                 window=int(window))}
+
+
+def _select_tests(flows: List[str], k: Optional[int] = None,
+                  threshold: Optional[float] = None,
+                  history_path: Optional[str] = None,
+                  window: int = 10) -> Dict[str, Any]:
+    """Adapter: pick the riskiest flows to run (top-k / threshold)."""
+    from je_auto_control.utils.test_select import select_flows
+    return {"selected": select_flows(
+        flows, k=k, threshold=threshold, history_path=history_path,
+        window=int(window))}
+
+
 class Executor:
     """
     Executor
@@ -2509,6 +2551,10 @@ def __init__(self):
             "AC_queue_complete": _queue_complete,
             "AC_queue_fail": _queue_fail,
             "AC_queue_stats": _queue_stats,
+            "AC_generate_data": _generate_data,
+            "AC_mcp_manifest": _mcp_manifest,
+            "AC_rank_tests": _rank_tests,
+            "AC_select_tests": _select_tests,
             "AC_a11y_record_start": _a11y_record_start,
             "AC_a11y_record_stop": _a11y_record_stop,
             "AC_a11y_record_events": _a11y_record_events,

je_auto_control/utils/mcp_registry/__init__.py

@@ -0,0 +1,6 @@
+"""MCP registry server.json manifest generation (discoverability)."""
+from je_auto_control.utils.mcp_registry.registry import (
+    build_server_manifest, write_server_manifest,
+)
+
+__all__ = ["build_server_manifest", "write_server_manifest"]