graphlemon: merge 39efb231-d574-446d-813d-9eabd50b5981

graphlemon
2 parents 9334f8d6 13a4350b
Showing 6 changed files with 372 additions and 12 deletions
frontend/app.js
frontend/test_run_detail_smoke.js
src/backend_server.py
src/test_backend_server.py
src/test_work_graph_run_query.py
src/work_graph_run_query.py
@@ -104,6 +104,11 @@ const ARTIFACT_DETAIL_RE = /^#\/artifacts\/([^/]+)$/;
 // reaching renderTaskDetail/mountTaskDetail with a bogus org_node that would
 // otherwise need its own defensive checks further down.
 const ORG_NODE_DETAIL_RE = /^#\/tasks\/([^/]+)\/org\/(planner|coder|validator)$/;
+// v_run_detail (U-046, 2026-08-06 SDD): "运行历史" (`#gc_runs_list`)
+// previously had nowhere real to send a click -- it just called
+// onSwitchToTask to switch v_graph_console's own task selector. This is the
+// actual run-detail page/route that closes that gap.
+const RUN_DETAIL_RE = /^#\/runs\/([^/]+)$/;
 let currentRouteCleanup = null;
 function registerRouteCleanup(fn) {
@@ -124,6 +129,7 @@ function render() {
     const orgNodeMatch = hash.match(ORG_NODE_DETAIL_RE);
     const skillMatch = hash.match(SKILL_AUDIT_RE);
     const artifactMatch = hash.match(ARTIFACT_DETAIL_RE);
+    const runMatch = hash.match(RUN_DETAIL_RE);
     if (orgNodeMatch) {
       // Reuses v_task_detail itself (same render/mount) with the org_node
       // carried through as a second arg -- no separate view/route handler.
@@ -140,6 +146,9 @@ function render() {
       // with the artifact_id pre-selected -- no separate detail view.
       route = { render: renderArtifactBrowser, mount: mountArtifactBrowser };
       routeArgs = [artifactMatch[1]];
+    } else if (runMatch) {
+      route = { render: renderRunDetail, mount: mountRunDetail };
+      routeArgs = [runMatch[1]];
     }
   }
   if (!route) route = { render: () => "<h1>Not found</h1><p>" + hash + "</p>" };
@@ -1468,6 +1477,122 @@ function mountTaskDetail(taskId, initialOrgNode) {
   });
 }
+// ---- v_run_detail (`#/runs/:run_id`, U-046 2026-08-06 SDD): single Work
+// Graph Run detail -- every work_node_execution the run contains, each with
+// its own log (live c_ws_log_tail while running, static c_execution_log_get
+// once terminal). Reuses renderGraphVisualizerPanel/mountLogViewer
+// unmodified, same as v_task_detail's todo tree above; the one DOM-id
+// contract mountLogViewer relies on (`#todo-detail-<node_exec_id>
+// .log-autoscroll`) is reproduced here via renderRunNodeListHtml so the
+// existing auto-scroll checkbox keeps working without touching
+// mountLogViewer itself.
+
+function renderRunNodeListHtml(run) {
+  if (!run.nodes.length) return '<div class="hint">该运行尚无节点</div>';
+  const items = run.nodes.map((wn) => {
+    const status = wn.status;
+    return (
+      '<li class="status-' + status + '">'
+      + '<div class="todo-row" data-node-exec-id="' + wn.node_exec_id + '">'
+      + '<span class="dot"></span>'
+      + '<span class="todo-id">' + escapeHtml(wn.node_id) + "</span>"
+      + '<span class="todo-status">' + escapeHtml(status) + " · " + escapeHtml(wn.subagent_type)
+      + " · retry_count=" + wn.retry_count + " · node_exec_id=" + wn.node_exec_id + "</span>"
+      + "</div>"
+      + '<div class="todo-detail" id="todo-detail-' + wn.node_exec_id + '" data-status="' + status + '">'
+      + (wn.reflection_notes ? "reflection_notes: " + escapeHtml(wn.reflection_notes) + "<br>" : "")
+      + '<div class="log-viewer-slot" id="log-slot-' + wn.node_exec_id + '"></div>'
+      + "</div>"
+      + "</li>"
+    );
+  }).join("");
+  return '<ul class="todo-tree">' + items + "</ul>";
+}
+
+function renderRunDetail(runId) {
+  if (!getCurrentProjectId()) return renderNoProjectSelected();
+  return `
+    <div class="page page-wide">
+      <div class="breadcrumb"><a href="#/graph">Graph 控制台</a> / run ${escapeHtml(runId)}</div>
+      <div id="rd_content"><div class="hint">加载中…</div></div>
+    </div>
+  `;
+}
+
+function mountRunDetail(runId) {
+  const projectId = getCurrentProjectId();
+  if (!projectId) return;
+  const content = document.getElementById("rd_content");
+
+  // Same "track open log viewers, close them on re-render/route-leave"
+  // pattern as mountTaskDetail/mountGraphConsole above.
+  const activeLogViewers = new Map();
+  function closeAllLogViewers() {
+    for (const close of activeLogViewers.values()) close();
+    activeLogViewers.clear();
+  }
+
+  // The event socket needs task_id, which we only learn from the run's own
+  // response -- opened lazily on the first successful render, not before.
+  let socket = null;
+
+  function renderDetail(run) {
+    closeAllLogViewers();
+    content.innerHTML = (
+      "<h1>" + escapeHtml(run.org_node) + " Work Graph Run</h1>"
+      + '<div class="meta-row">'
+      + "<div><b>RUN_ID</b>" + escapeHtml(run.run_id) + "</div>"
+      + "<div><b>STATUS</b>" + escapeHtml(run.status) + "</div>"
+      + "<div><b>STARTED_AT</b>" + escapeHtml(run.started_at || "") + "</div>"
+      + "<div><b>COMPLETED_AT</b>" + escapeHtml(run.completed_at || "") + "</div>"
+      + '<div><b>TASK</b><a href="#/tasks/' + encodeURIComponent(run.task_id) + '">' + escapeHtml(run.task_id) + "</a></div>"
+      + "</div>"
+      + renderGraphVisualizerPanel(run)
+      + '<div class="tooltip" id="graph-tooltip"></div>'
+      + '<div class="panel"><h2>节点列表</h2>' + renderRunNodeListHtml(run) + "</div>"
+    );
+
+    for (const row of content.querySelectorAll(".todo-row[data-node-exec-id]")) {
+      row.addEventListener("click", () => {
+        const nodeExecId = row.dataset.nodeExecId;
+        if (!nodeExecId) return;
+        const detail = document.getElementById("todo-detail-" + nodeExecId);
+        if (!detail) return;
+        detail.classList.toggle("open");
+        if (detail.classList.contains("open") && !activeLogViewers.has(nodeExecId)) {
+          const slot = document.getElementById("log-slot-" + nodeExecId);
+          slot.innerHTML = renderLogViewerHtml(nodeExecId);
+          activeLogViewers.set(nodeExecId, mountLogViewer(nodeExecId, detail.dataset.status));
+        }
+      });
+    }
+    mountGraphVisualizerTooltips(content);
+
+    if (!socket) {
+      socket = connectEventSocket(projectId, run.task_id, {
+        onEvent: () => refresh(),
+        onStatusChange: () => {},
+      });
+    }
+  }
+
+  async function refresh() {
+    const resp = await fetch("/api/work-graph-runs/" + encodeURIComponent(runId));
+    if (resp.status === 404) {
+      content.innerHTML = '<div class="error-banner" style="display:block;">运行不存在</div>';
+      return;
+    }
+    renderDetail(await resp.json());
+  }
+
+  refresh();
+
+  registerRouteCleanup(() => {
+    if (socket) socket.close();
+    closeAllLogViewers();
+  });
+}
+
 // ---- v_graph_console (`#/graph`, 2026-08-03 SDD): Org/Work Graph
 // visualization + Planner output overview. FR-1: independent Org Graph
 // topology (not `org-strip`, see renderOrgGraphTopology below). FR-2/3:
@@ -1525,9 +1650,11 @@ function renderArtifactSummaryList(artifacts, types, emptyText) {
 // #/artifacts / #/tasks with no filter pre-applied -- v_artifact_browser
 // and v_task_board don't parse URL query params today, and adding that is
 // out of scope for this page (SDD §2.5's explicit, recorded trade-off).
-// Graph run rows are the one exception: clicking a row switches this same
-// page's own task switcher, since there's no separate "run detail" route.
-async function mountArtifactsOverviewPanel(container, projectId, { onSwitchToTask }) {
+// Graph run rows: U-046 fixed the previous shortcut (switching this same
+// page's own task switcher via onSwitchToTask, because there was no
+// separate "run detail" route) to navigate to the real v_run_detail page
+// (#/runs/:run_id) instead.
+async function mountArtifactsOverviewPanel(container, projectId) {
   container.innerHTML = '<div class="panel"><h2>Planner 产出总览</h2><div class="hint">加载中…</div></div>';
   const [artifacts, runs, tasks] = await Promise.all([
     fetch("/api/artifacts?project_id=" + encodeURIComponent(projectId)).then((r) => r.json()),
@@ -1540,7 +1667,7 @@ async function mountArtifactsOverviewPanel(container, projectId, { onSwitchToTas
   const runsHtml = runs.length === 0 ? '<div class="hint">暂无 Graph 运行记录</div>' : (
     '<ul class="simple-list" id="gc_runs_list">' + runs.slice(0, 10).map((r) =>
-      '<li data-run-task-id="' + r.task_id + '" style="cursor:pointer;"><span>' + escapeHtml(r.org_node) + " · run "
+      '<li data-run-id="' + r.run_id + '" style="cursor:pointer;"><span>' + escapeHtml(r.org_node) + " · run "
       + r.run_id.slice(0, 8) + "</span><span>" + escapeHtml(r.status) + " · " + r.node_count + " 节点</span></li>"
     ).join("") + "</ul>"
   );
@@ -1564,8 +1691,10 @@ async function mountArtifactsOverviewPanel(container, projectId, { onSwitchToTas
   const runsListEl = document.getElementById("gc_runs_list");
   if (runsListEl) {
-    for (const li of runsListEl.querySelectorAll("li[data-run-task-id]")) {
-      li.addEventListener("click", () => onSwitchToTask(li.dataset.runTaskId));
+    for (const li of runsListEl.querySelectorAll("li[data-run-id]")) {
+      li.addEventListener("click", () => {
+        window.location.hash = "#/runs/" + encodeURIComponent(li.dataset.runId);
+      });
     }
   }
@@ -1639,12 +1768,7 @@ function mountGraphConsole() {
   }
   function refreshArtifactsOverview() {
-    mountArtifactsOverviewPanel(artifactsOverviewEl, projectId, {
-      onSwitchToTask: (taskId) => {
-        taskSelect.value = taskId;
-        loadTaskAndRender(taskId);
-      },
-    });
+    mountArtifactsOverviewPanel(artifactsOverviewEl, projectId);
   }
   async function loadTaskAndRender(taskId) {
+// v_run_detail (`#/runs/:run_id`, U-046 2026-08-06 SDD) smoke test. Same
+// no-framework/no-fixture convention as frontend/test_graph_console_smoke.js
+// and frontend/test_org_click_through_smoke.js (see
+// .claude/agent-memory/validator/feedback_no_js_test_framework.md) -- loads
+// app.js in a bare vm.Context and asserts on generated HTML strings / regex
+// matches for the new top-level pieces:
+//   - RUN_DETAIL_RE: the new #/runs/:run_id route.
+//   - renderRunNodeListHtml: the run-detail node list, including the
+//     todo-detail-<node_exec_id> DOM-id contract mountLogViewer relies on
+//     (`#todo-detail-<id> .log-autoscroll`) and the empty-run case.
+//   - mountArtifactsOverviewPanel's now-fixed click target: `#gc_runs_list`
+//     rows carry data-run-id (not the old data-run-task-id), so a real
+//     click can build "#/runs/" + run_id instead of only ever switching
+//     the task selector.
+//
+// Run: node frontend/test_run_detail_smoke.js
+
+"use strict";
+const fs = require("fs");
+const path = require("path");
+const vm = require("vm");
+const assert = require("assert");
+
+function makeFakeDiv() {
+  let text = "";
+  return {
+    set textContent(v) { text = v == null ? "" : String(v); },
+    get textContent() { return text; },
+    get innerHTML() {
+      return text.replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;").replace(/"/g, "&quot;");
+    },
+  };
+}
+
+const sandbox = {
+  document: {
+    createElement: () => makeFakeDiv(),
+    getElementById: () => null,
+    querySelectorAll: () => [],
+    addEventListener: () => {},
+  },
+  window: {
+    location: { protocol: "http:", host: "localhost", hash: "" },
+    addEventListener: () => {},
+  },
+  localStorage: { getItem: () => null, setItem: () => {} },
+  console,
+};
+vm.createContext(sandbox);
+
+const src = fs.readFileSync(path.join(__dirname, "app.js"), "utf8");
+// Top-level `const` regexes stay lexically scoped to the script and never
+// reach the sandbox object (same Annex B quirk the other smoke tests
+// already document) -- re-export what this test needs.
+const augmentedSrc = src + "\nglobalThis.RUN_DETAIL_RE = RUN_DETAIL_RE;\n";
+new vm.Script(augmentedSrc, { filename: "app.js" }).runInContext(sandbox);
+
+let assertionCount = 0;
+function check(condition, message) {
+  assertionCount++;
+  assert.ok(condition, message);
+}
+
+// ---- 1. RUN_DETAIL_RE: happy / boundary / failure ----
+{
+  const legal = "#/runs/run-abc-123".match(sandbox.RUN_DETAIL_RE);
+  check(legal, "legal run_id must match");
+  check(legal[1] === "run-abc-123", "captured group must be the run_id");
+
+  check(!"#/runs/".match(sandbox.RUN_DETAIL_RE), "missing run_id segment must not match");
+  check(!"#/runs/abc/extra".match(sandbox.RUN_DETAIL_RE), "extra path segments must not match");
+  check(!"#/tasks/abc".match(sandbox.RUN_DETAIL_RE), "an unrelated route must not match");
+}
+
+// ---- 2. renderRunNodeListHtml: node list + todo-detail-<id> DOM contract ----
+{
+  const run = {
+    run_id: "run-1", org_node: "coder", status: "running",
+    nodes: [
+      {
+        node_exec_id: "exec-1", node_id: "n1", subagent_type: "coder", status: "running",
+        retry_count: 0, reflection_notes: null,
+      },
+      {
+        node_exec_id: "exec-2", node_id: "n2", subagent_type: "coder", status: "done",
+        retry_count: 1, reflection_notes: "retried once after a transient error",
+      },
+    ],
+  };
+  const html = sandbox.renderRunNodeListHtml(run);
+  check(html.includes('data-node-exec-id="exec-1"'), "each row must carry its node_exec_id for the click handler");
+  check(html.includes('id="todo-detail-exec-1"'), "must reproduce the todo-detail-<id> id mountLogViewer's querySelector needs");
+  check(html.includes('data-status="running"'), "todo-detail wrapper must carry the node's live status");
+  check(html.includes('id="log-slot-exec-2"'), "must reserve a log-viewer-slot per node, same as v_task_detail's todo tree");
+  check(html.includes("retried once after a transient error"), "reflection_notes must be shown when present");
+  check(!html.includes("reflection_notes: <br>"), "no stray reflection_notes line when it's null (exec-1)");
+
+  const emptyHtml = sandbox.renderRunNodeListHtml({ nodes: [] });
+  check(emptyHtml.includes("该运行尚无节点"), "a run with zero nodes must render an explicit empty state, not blank/undefined");
+}
+
+// ---- 3. gc_runs_list rows: run-detail click target (2026-08-06 fix) ----
+{
+  // mountArtifactsOverviewPanel itself does an async fetch and can't run in
+  // this sandbox (no `fetch`), but the row-id contract it depends on is a
+  // pure string-building concern worth pinning directly: replicate the
+  // exact <li> markup mountArtifactsOverviewPanel emits for one run and
+  // confirm it carries data-run-id (not the removed data-run-task-id).
+  const escapeHtml = sandbox.escapeHtml;
+  const r = { org_node: "coder", run_id: "run-1234567890", status: "done", node_count: 3 };
+  const li = '<li data-run-id="' + r.run_id + '" style="cursor:pointer;"><span>' + escapeHtml(r.org_node) + " · run "
+    + r.run_id.slice(0, 8) + "</span><span>" + escapeHtml(r.status) + " · " + r.node_count + " 节点</span></li>";
+  check(li.includes('data-run-id="run-1234567890"'), "run row must carry the real run_id for #/runs/:run_id navigation");
+  check(!li.includes("data-run-task-id"), "the old task-switcher click target must be gone");
+}
+
+console.log("frontend/test_run_detail_smoke.js: all " + assertionCount + " assertions passed");
@@ -130,6 +130,7 @@ _LOG_SUFFIX = &quot;/log&quot;
 _ARTIFACTS_PREFIX = "/api/artifacts/"
 _SKILLS_PREFIX = "/api/skills/"
 _DEPLOYMENTS_PREFIX = "/api/deployments/"
+_WORK_GRAPH_RUNS_PREFIX = "/api/work-graph-runs/"
 class _Handler(BaseHTTPRequestHandler):
@@ -180,6 +181,11 @@ class _Handler(BaseHTTPRequestHandler):
         if parsed.path == "/api/work-graph-runs":
             self._handle_work_graph_run_list(parsed.query)
             return
+        if parsed.path.startswith(_WORK_GRAPH_RUNS_PREFIX):
+            run_id = parsed.path[len(_WORK_GRAPH_RUNS_PREFIX):]
+            if run_id and "/" not in run_id:
+                self._handle_work_graph_run_get(run_id)
+                return
         if parsed.path.startswith(_ARTIFACTS_PREFIX):
             artifact_id = parsed.path[len(_ARTIFACTS_PREFIX):]
             if artifact_id and "/" not in artifact_id:
@@ -666,6 +672,20 @@ class _Handler(BaseHTTPRequestHandler):
         else:
             self._send_json(200, result)
+    def _handle_work_graph_run_get(self, run_id):
+        """c_work_graph_run_get (U-046, GET /api/work-graph-runs/:run_id):
+        v_run_detail's single-run lookup -- the list endpoint above has no
+        per-run detail shape, and v_run_detail needs the full node list
+        (same shape as task_query.get_task's embedded work_graph_runs[i])
+        to reuse renderGraphVisualizerPanel/mountLogViewer unmodified."""
+        try:
+            with self.server.conn_lock:
+                result = work_graph_run_query.get_work_graph_run(self.server.conn, run_id)
+        except work_graph_run_query.WorkGraphRunNotFoundError:
+            self._send_json(404, {"error": "NOT_FOUND"})
+        else:
+            self._send_json(200, result)
+
     def _handle_artifact_get(self, artifact_id):
         try:
             with self.server.conn_lock:
@@ -1354,6 +1354,19 @@ class WorkGraphRunListRouteTest(unittest.TestCase):
         self.assertEqual(status, 200)
         self.assertEqual(body, [])
+    def test_get_returns_run_with_nodes(self):
+        status, body = self._get("/api/work-graph-runs/run-1")
+        self.assertEqual(status, 200)
+        self.assertEqual(body["run_id"], "run-1")
+        self.assertEqual(body["task_id"], "task-1")
+        self.assertEqual(body["org_node"], "coder")
+        self.assertEqual(body["nodes"], [])
+
+    def test_get_nonexistent_run_returns_404(self):
+        status, body = self._get("/api/work-graph-runs/no-such-run")
+        self.assertEqual(status, 404)
+        self.assertEqual(body["error"], "NOT_FOUND")
+
 class DeploymentRouteTest(unittest.TestCase):
     """U-043: c_deployment_list/c_deployment_current/c_deployment_get for
@@ -118,5 +118,33 @@ class ListWorkGraphRunsTest(WorkGraphRunQueryTestBase):
         self.assertEqual(result, [])
+class GetWorkGraphRunTest(WorkGraphRunQueryTestBase):
+    """U-046: c_work_graph_run_get -- v_run_detail's single-run lookup."""
+
+    def test_returns_run_with_embedded_nodes(self):
+        self._insert_run("run-1", "coder", status="running", node_count=2)
+        result = work_graph_run_query.get_work_graph_run(self.conn, "run-1")
+        self.assertEqual(result["run_id"], "run-1")
+        self.assertEqual(result["task_id"], "task-1")
+        self.assertEqual(result["org_node"], "coder")
+        self.assertEqual(result["status"], "running")
+        self.assertEqual(result["graph_manifest"], {})
+        self.assertEqual(len(result["nodes"]), 2)
+        self.assertEqual(
+            {n["node_exec_id"] for n in result["nodes"]},
+            {"run-1-node-0", "run-1-node-1"},
+        )
+        self.assertEqual(result["nodes"][0]["status"], "done")
+
+    def test_nonexistent_run_raises(self):
+        with self.assertRaises(work_graph_run_query.WorkGraphRunNotFoundError):
+            work_graph_run_query.get_work_graph_run(self.conn, "no-such-run")
+
+    def test_run_with_no_nodes_returns_empty_list(self):
+        self._insert_run("run-1", "planner", node_count=0)
+        result = work_graph_run_query.get_work_graph_run(self.conn, "run-1")
+        self.assertEqual(result["nodes"], [])
+
+
 if __name__ == "__main__":
     unittest.main()
@@ -7,8 +7,21 @@ codebase that lists `work_graph_run` rows across all tasks in a project
 rather than for one `task_id` at a time, which is why it lives in its own
 module instead of `task_query.py` (that module's own docstring frames it
 as "built around a single task_id").
+
+U-046 (2026-08-06 SDD, "运行历史点击跳转错误 + 缺运行详情页"): adds
+`get_work_graph_run`, a single-`run_id` lookup for the new `v_run_detail`
+page (`#/runs/:run_id`). This deliberately mirrors the shape of
+`task_query.get_task`'s embedded `work_graph_runs[i]` entries (same keys,
+same per-node field list via `_nodes_for_run`) so the frontend can reuse
+`renderGraphVisualizerPanel`/`mountLogViewer` against this response without
+any remapping -- it's a separate query (by `run_id` instead of `task_id`)
+rather than a call into `task_query._work_nodes_for_run` because that
+helper is private to its own module and this module already owns
+`work_graph_run` reads.
 """
+import json
+
 _VALID_ORG_NODES = ("planner", "coder", "validator")
@@ -51,3 +64,48 @@ def list_work_graph_runs(conn, project_id, org_node=None, limit=20):
         }
         for r in rows
     ]
+
+
+class WorkGraphRunNotFoundError(Exception):
+    """NOT_FOUND -- c_work_graph_run_get on a nonexistent run_id."""
+
+
+def get_work_graph_run(conn, run_id):
+    row = conn.execute(
+        "SELECT run_id, task_id, org_node, graph_manifest, status, started_at, completed_at "
+        "FROM work_graph_run WHERE run_id = ?",
+        (run_id,),
+    ).fetchone()
+    if row is None:
+        raise WorkGraphRunNotFoundError(run_id)
+    return {
+        "run_id": row[0], "task_id": row[1], "org_node": row[2],
+        "graph_manifest": json.loads(row[3]) if row[3] else None,
+        "status": row[4], "started_at": row[5], "completed_at": row[6],
+        "nodes": _nodes_for_run(conn, row[0]),
+    }
+
+
+def _nodes_for_run(conn, run_id):
+    # Same column list as task_query._work_nodes_for_run -- kept in lockstep
+    # on purpose (see module docstring) so v_run_detail can reuse
+    # renderGraphVisualizerPanel/mountLogViewer unmodified.
+    rows = conn.execute(
+        "SELECT node_exec_id, node_id, subagent_type, status, retry_count, input_payload, "
+        "output_payload, reflection_notes, git_worktree_ref, input_schema_ref, output_schema_ref, "
+        "quota_limit_type, quota_wait_started_at "
+        "FROM work_node_execution WHERE run_id = ? ORDER BY node_exec_id",
+        (run_id,),
+    ).fetchall()
+    return [
+        {
+            "node_exec_id": n[0], "node_id": n[1], "subagent_type": n[2], "status": n[3],
+            "retry_count": n[4],
+            "input_payload": json.loads(n[5]) if n[5] else None,
+            "output_payload": json.loads(n[6]) if n[6] else None,
+            "reflection_notes": n[7], "git_worktree_ref": n[8],
+            "input_schema_ref": n[9], "output_schema_ref": n[10],
+            "quota_limit_type": n[11], "quota_wait_started_at": n[12],
+        }
+        for n in rows
+    ]