Bring MCP server up to date: 9 -> 19 tools + dict-shape lint fix

The MCP server was last touched when only nine helpers existed; ~30
new pure-function modules have landed since (action_formatter,
md_authoring, sel_to_pw, pom_codegen, pii_scanner, failure_cluster,
a11y_diff, locator_strength's batch scorer). This wires the most
useful 10 of them into ``build_default_tools()`` so MCP-aware clients
can call them directly.

New tools:
- webrunner_format_actions
- webrunner_parse_markdown
- webrunner_translate_actions_to_playwright
- webrunner_translate_python_to_playwright
- webrunner_pom_from_html
- webrunner_scan_pii
- webrunner_redact_pii
- webrunner_cluster_failures
- webrunner_a11y_diff
- webrunner_score_action_locators

Bug: _tool_lint_action used dataclass-style attribute access (f.level,
f.message) but ``lint_action`` actually returns ``List[Dict[str, Any]]``
keyed on rule / severity / message / location. Pass the list through
verbatim so MCP clients see the same shape the Python API exposes.

Test: TestNewTools covers each new tool end-to-end through the MCP
server. TestDefaultTools.test_full_default_tool_surface freezes the
19-tool list so accidental removals fail loudly in review.

README + EN/ZH Sphinx docs grouped the now-larger tool list by purpose
(authoring / codegen / quality / security / reporting / infra) so
operators can scan the surface at a glance.

Tests: 1230 -> 1241.

Author: JE-Chen

README.md

@@ -687,11 +687,26 @@ WebRunner ships a [Model Context Protocol](https://modelcontextprotocol.io/) ser
 python -m je_web_runner.mcp_server
 ```
 
-The default tool list exposes:
+The default tool list (19 tools) exposes:
 
-- `webrunner_lint_action`, `webrunner_locator_strength`
-- `webrunner_render_template`, `webrunner_compute_trend`
-- `webrunner_validate_response`, `webrunner_summary_markdown`
+Action JSON authoring & linting:
+- `webrunner_lint_action`, `webrunner_score_action_locators`, `webrunner_locator_strength`
+- `webrunner_format_actions`, `webrunner_parse_markdown`, `webrunner_render_template`
+- `webrunner_translate_actions_to_playwright`, `webrunner_translate_python_to_playwright`
+
+Code generation:
+- `webrunner_pom_from_html`
+
+Quality / triage:
+- `webrunner_a11y_diff`, `webrunner_cluster_failures`, `webrunner_compute_trend`
+
+Security & privacy:
+- `webrunner_scan_pii`, `webrunner_redact_pii`
+
+Reporting & contract:
+- `webrunner_summary_markdown`, `webrunner_validate_response`
+
+Sharding & infra:
 - `webrunner_diff_shard`, `webrunner_render_k8s`, `webrunner_partition_shard`
 
 ```python

docs/source/Eng/doc/extended_features/extended_features_doc.rst

@@ -380,11 +380,22 @@ drive it over JSON-RPC stdio:
 
    python -m je_web_runner.mcp_server
 
-Default tools registered: ``webrunner_lint_action``,
-``webrunner_locator_strength``, ``webrunner_render_template``,
-``webrunner_compute_trend``, ``webrunner_validate_response``,
-``webrunner_summary_markdown``, ``webrunner_diff_shard``,
-``webrunner_render_k8s``, ``webrunner_partition_shard``.
+Default tools registered (19 in total):
+
+* Action authoring & lint: ``webrunner_lint_action``,
+  ``webrunner_score_action_locators``, ``webrunner_locator_strength``,
+  ``webrunner_format_actions``, ``webrunner_parse_markdown``,
+  ``webrunner_render_template``,
+  ``webrunner_translate_actions_to_playwright``,
+  ``webrunner_translate_python_to_playwright``
+* Code generation: ``webrunner_pom_from_html``
+* Quality & triage: ``webrunner_a11y_diff``,
+  ``webrunner_cluster_failures``, ``webrunner_compute_trend``
+* Security: ``webrunner_scan_pii``, ``webrunner_redact_pii``
+* Reporting & contract: ``webrunner_summary_markdown``,
+  ``webrunner_validate_response``
+* Sharding / infra: ``webrunner_diff_shard``,
+  ``webrunner_render_k8s``, ``webrunner_partition_shard``
 
 Custom tools register via ``McpServer.register(Tool(...))``; the server
 implements MCP ``2024-11-05`` (``initialize`` / ``tools/list`` /

docs/source/Zh/doc/extended_features/extended_features_doc.rst

@@ -264,12 +264,25 @@ MCP server
 
    python -m je_web_runner.mcp_server
 
-預設工具：``webrunner_lint_action`` / ``webrunner_locator_strength`` /
-``webrunner_render_template`` / ``webrunner_compute_trend`` /
-``webrunner_validate_response`` / ``webrunner_summary_markdown`` /
-``webrunner_diff_shard`` / ``webrunner_render_k8s`` /
-``webrunner_partition_shard``。可透過 ``McpServer.register(Tool(...))``
-擴充自訂工具，協定版本 ``2024-11-05``。
+預設工具共 19 個，依用途分組：
+
+* Action 撰寫 / lint：``webrunner_lint_action`` /
+  ``webrunner_score_action_locators`` / ``webrunner_locator_strength`` /
+  ``webrunner_format_actions`` / ``webrunner_parse_markdown`` /
+  ``webrunner_render_template`` /
+  ``webrunner_translate_actions_to_playwright`` /
+  ``webrunner_translate_python_to_playwright``
+* 程式碼生成：``webrunner_pom_from_html``
+* 品質 / triage：``webrunner_a11y_diff`` / ``webrunner_cluster_failures``
+  / ``webrunner_compute_trend``
+* 安全 / 隱私：``webrunner_scan_pii`` / ``webrunner_redact_pii``
+* 報告 / contract：``webrunner_summary_markdown`` /
+  ``webrunner_validate_response``
+* Sharding / infra：``webrunner_diff_shard`` / ``webrunner_render_k8s``
+  / ``webrunner_partition_shard``
+
+可透過 ``McpServer.register(Tool(...))`` 自行擴充工具，協定版本
+``2024-11-05``。
 
 Action JSON LSP
 ===============

je_web_runner/mcp_server/server.py

@@ -145,8 +145,10 @@ def _tool_lint_action(arguments: Dict[str, Any]) -> Any:
     actions = arguments.get("actions")
     if not isinstance(actions, list):
         raise McpServerError("'actions' must be a list")
-    return [{"index": f.index, "level": f.level, "message": f.message,
-             "rule": f.rule} for f in lint_action(actions)]
+    # ``lint_action`` returns ``List[Dict[str, Any]]`` with ``rule`` /
+    # ``severity`` / ``message`` / ``location`` keys; pass through verbatim
+    # so MCP clients see the same shape the Python API exposes.
+    return list(lint_action(actions))
 
 
 def _tool_locator_strength(arguments: Dict[str, Any]) -> Any:
@@ -232,6 +234,127 @@ def _tool_partition(arguments: Dict[str, Any]) -> Any:
     )
 
 
+def _tool_format_actions(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.action_formatter.formatter import format_actions
+    actions = arguments.get("actions")
+    if not isinstance(actions, list):
+        raise McpServerError("'actions' must be a list")
+    return format_actions(actions, indent=int(arguments.get("indent", 2)))
+
+
+def _tool_parse_markdown(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.md_authoring.markdown_to_actions import parse_markdown
+    text = arguments.get("text")
+    if not isinstance(text, str):
+        raise McpServerError("'text' must be a string")
+    return parse_markdown(text)
+
+
+def _tool_translate_actions_to_playwright(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.sel_to_pw.translator import translate_action_list
+    actions = arguments.get("actions")
+    if not isinstance(actions, list):
+        raise McpServerError("'actions' must be a list")
+    return translate_action_list(actions)
+
+
+def _tool_translate_python_to_playwright(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.sel_to_pw.translator import translate_python_source
+    source = arguments.get("source")
+    if not isinstance(source, str):
+        raise McpServerError("'source' must be a string")
+    translations = translate_python_source(source)
+    return [
+        {"line": t.line, "original": t.original,
+         "translated": t.translated, "note": t.note}
+        for t in translations
+    ]
+
+
+def _tool_pom_from_html(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.pom_codegen.codegen import (
+        discover_elements_from_html,
+        render_pom_module,
+    )
+    html = arguments.get("html")
+    if not isinstance(html, str):
+        raise McpServerError("'html' must be a string")
+    elements = discover_elements_from_html(html)
+    class_name = str(arguments.get("class_name", "WebRunnerPage"))
+    return {
+        "module": render_pom_module(elements, class_name=class_name),
+        "elements": [
+            {"name": e.name, "strategy": e.strategy,
+             "value": e.value, "tag": e.tag, "source": e.source}
+            for e in elements
+        ],
+    }
+
+
+def _tool_scan_pii(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.pii_scanner.scanner import scan_text
+    text = arguments.get("text")
+    if not isinstance(text, str):
+        raise McpServerError("'text' must be a string")
+    categories = arguments.get("categories")
+    findings = scan_text(text, categories=categories)
+    return [
+        {"category": f.category, "start": f.start,
+         "end": f.end, "redacted": f.redacted}
+        for f in findings
+    ]
+
+
+def _tool_redact_pii(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.pii_scanner.scanner import redact_text
+    text = arguments.get("text")
+    if not isinstance(text, str):
+        raise McpServerError("'text' must be a string")
+    return redact_text(
+        text,
+        replacement=str(arguments.get("replacement", "[REDACTED]")),
+        categories=arguments.get("categories"),
+    )
+
+
+def _tool_cluster_failures(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.failure_cluster.clustering import (
+        cluster_failures,
+        cluster_summary,
+    )
+    failures = arguments.get("failures")
+    if not isinstance(failures, list):
+        raise McpServerError("'failures' must be a list")
+    top_n = arguments.get("top_n")
+    if top_n is not None:
+        top_n = int(top_n)
+    clusters = cluster_failures(failures, top_n=top_n)
+    return cluster_summary(clusters)
+
+
+def _tool_a11y_diff(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.accessibility.a11y_diff import diff_violations
+    baseline = arguments.get("baseline")
+    current = arguments.get("current")
+    if not isinstance(baseline, list) or not isinstance(current, list):
+        raise McpServerError("'baseline' and 'current' must be lists")
+    diff = diff_violations(baseline, current)
+    return {
+        "added": diff.added,
+        "resolved": diff.resolved,
+        "persisting": diff.persisting,
+        "regressed": diff.regressed,
+    }
+
+
+def _tool_score_action_locators(arguments: Dict[str, Any]) -> Any:
+    from je_web_runner.utils.linter.locator_strength import score_action_locators
+    actions = arguments.get("actions")
+    if not isinstance(actions, list):
+        raise McpServerError("'actions' must be a list")
+    return list(score_action_locators(actions))
+
+
 def build_default_tools() -> List[Tool]:
     """Construct the default tool list shipped with the server."""
     return [
@@ -351,6 +474,146 @@ def build_default_tools() -> List[Tool]:
             },
             handler=_tool_partition,
         ),
+        Tool(
+            name="webrunner_format_actions",
+            description="Format an action JSON list with canonical kwarg order.",
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "actions": {"type": "array"},
+                    "indent": {"type": "integer"},
+                },
+                "required": ["actions"],
+            },
+            handler=_tool_format_actions,
+        ),
+        Tool(
+            name="webrunner_parse_markdown",
+            description="Transpile a Markdown bullet list into a WR_* action list.",
+            input_schema={
+                "type": "object",
+                "properties": {"text": {"type": "string"}},
+                "required": ["text"],
+            },
+            handler=_tool_parse_markdown,
+        ),
+        Tool(
+            name="webrunner_translate_actions_to_playwright",
+            description=(
+                "Rewrite a WR_* action list to its WR_pw_* Playwright"
+                " equivalent (drops WR_implicitly_wait)."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {"actions": {"type": "array"}},
+                "required": ["actions"],
+            },
+            handler=_tool_translate_actions_to_playwright,
+        ),
+        Tool(
+            name="webrunner_translate_python_to_playwright",
+            description=(
+                "Static translator: rewrites Selenium-style Python source"
+                " into Playwright equivalents; returns per-line diffs."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {"source": {"type": "string"}},
+                "required": ["source"],
+            },
+            handler=_tool_translate_python_to_playwright,
+        ),
+        Tool(
+            name="webrunner_pom_from_html",
+            description=(
+                "Discover [data-testid] / id / form fields in HTML and"
+                " render a Python Page Object module."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "html": {"type": "string"},
+                    "class_name": {"type": "string"},
+                },
+                "required": ["html"],
+            },
+            handler=_tool_pom_from_html,
+        ),
+        Tool(
+            name="webrunner_scan_pii",
+            description=(
+                "Scan text for PII (email / phone / Luhn-card / SSN /"
+                " ROC ID / IPv4); returns category + redacted preview."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "text": {"type": "string"},
+                    "categories": {"type": "array", "items": {"type": "string"}},
+                },
+                "required": ["text"],
+            },
+            handler=_tool_scan_pii,
+        ),
+        Tool(
+            name="webrunner_redact_pii",
+            description="Replace each detected PII match with a sentinel string.",
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "text": {"type": "string"},
+                    "replacement": {"type": "string"},
+                    "categories": {"type": "array", "items": {"type": "string"}},
+                },
+                "required": ["text"],
+            },
+            handler=_tool_redact_pii,
+        ),
+        Tool(
+            name="webrunner_cluster_failures",
+            description=(
+                "Group failures by normalised error signature; returns"
+                " top buckets sorted by count."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "failures": {"type": "array"},
+                    "top_n": {"type": "integer"},
+                },
+                "required": ["failures"],
+            },
+            handler=_tool_cluster_failures,
+        ),
+        Tool(
+            name="webrunner_a11y_diff",
+            description=(
+                "Diff two axe-core ``violations`` arrays and bucket the"
+                " findings into added / resolved / persisting."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {
+                    "baseline": {"type": "array"},
+                    "current": {"type": "array"},
+                },
+                "required": ["baseline", "current"],
+            },
+            handler=_tool_a11y_diff,
+        ),
+        Tool(
+            name="webrunner_score_action_locators",
+            description=(
+                "Score every locator referenced by an action JSON list on"
+                " a 0–100 scale; lower = more fragile."
+            ),
+            input_schema={
+                "type": "object",
+                "properties": {"actions": {"type": "array"}},
+                "required": ["actions"],
+            },
+            handler=_tool_score_action_locators,
+        ),
     ]