feat: add agent session route

.compozy/tasks/qa-rounds/reviews-001/_meta.md

@@ -1,11 +1,11 @@
 ---
 provider: coderabbit
-pr: "78"
+pr: "80"
 round: 1
-created_at: 2026-04-26T20:29:02.311014Z
+created_at: 2026-04-27T00:34:22.802061Z
 ---
 
 ## Summary
-- Total: 33
+- Total: 13
 - Resolved: 0
-- Unresolved: 33
+- Unresolved: 13

.compozy/tasks/qa-rounds/reviews-001/issue_001.md

@@ -1,23 +1,26 @@
 ---
 status: resolved
-file: internal/api/core/agent_identity.go
-line: 148
+file: web/src/components/app-sidebar.test.tsx
+line: 263
 severity: nitpick
 author: coderabbitai[bot]
-provider_ref: review:4177411700,nitpick_hash:ee846dd868b1
-review_hash: ee846dd868b1
-source_review_id: "4177411700"
-source_review_submitted_at: "2026-04-26T20:24:27Z"
+provider_ref: review:4177569108,nitpick_hash:ad1c523077c7
+review_hash: ad1c523077c7
+source_review_id: "4177569108"
+source_review_submitted_at: "2026-04-26T22:35:58Z"
 ---
 
-# Issue 001: Consider wrapping the error with context.
+# Issue 001: Make this route mock param-aware.
 ## Review Comment
 
-Line 158 returns the error from `ResolveCoordinatorConfig` without wrapping. Adding context would help trace failures.
-
-As per coding guidelines: "Use explicit error returns with wrapped context: `fmt.Errorf("context: %w", err)` in Go".
+The current stub keys only on `to`, so setting `matchedRouteFuzzy["/agents/$name"] = true` would mark every agent row active. That makes this assertion non-discriminating for the `name` param handling you want to cover.
 
 ## Triage
 
 - Decision: `VALID`
-- Notes: `agentCoordinatorConfigPayload` returns the raw `ResolveCoordinatorConfig` error. This violates the local wrapped-error convention and loses call-site context. Fix by wrapping the resolver failure with `fmt.Errorf("resolve coordinator config: %w", err)`.
+- Notes:
+  - The `useMatchRoute` test stub only keys route activity by `to`, while `AppSidebar` calls `matchRoute({ to: "/agents/$name", params: { name: agent.name }, fuzzy: true })`.
+  - Setting a fuzzy match for `"/agents/$name"` would therefore mark every agent row active and fail to prove that the `name` param is respected.
+  - Fix by making the test mock key on `to + params` and updating the active-row assertion to render multiple agents, with only the matching param active.
+  - Resolution: implemented param-aware route match keys and asserted that only the matching agent row is active.
+  - Verification: targeted Vitest passed; `make verify` passed.

.compozy/tasks/qa-rounds/reviews-001/issue_002.md

@@ -1,39 +1,54 @@
 ---
 status: resolved
-file: internal/api/udsapi/agent_channels_test.go
-line: 116
+file: web/src/routes/_app/agents.$name.sessions.$id.tsx
+line: 136
 author: coderabbitai[bot]
-provider_ref: thread:PRRT_kwDOR5y4QM59r7vG,comment:PRRC_kwDOR5y4QM67Z0NA
+provider_ref: thread:PRRT_kwDOR5y4QM59sdE9,comment:PRRC_kwDOR5y4QM67ae48
 ---
 
-# Issue 002: _⚠️ Potential issue_ | _🟠 Major_
+# Issue 002: _⚠️ Potential issue_ | _🟡 Minor_
 ## Review Comment
 
-_⚠️ Potential issue_ | _🟠 Major_
+_⚠️ Potential issue_ | _🟡 Minor_
 
-**Wrap this test case in `t.Run("Should ...")` to match required test structure.**
+**Use canonical agent name for post-delete navigation.**
 
-This new test is currently a direct top-level body without the required subtest naming pattern.
+`onDeleteSuccess` currently routes with the URL param `name`, which can be stale if the path is manually edited or mismatched. Use the resolved session agent name for consistent redirect behavior.
 
+<details>
+<summary>🐛 Suggested fix</summary>
 
-As per coding guidelines, "MUST use t.Run("Should...") pattern for ALL test cases".
+```diff
+   const workspaceName = workspaces?.find(workspace => workspace.id === session.workspace_id)?.name;
++  const resolvedAgentName = session.agent_name ?? name;
+@@
+       <SessionPageContent
+-        agentName={session.agent_name ?? name}
++        agentName={resolvedAgentName}
+         sessionId={id}
+         session={session}
+         workspaceName={workspaceName}
+         onDeleteSuccess={() => {
+-          void navigate({ to: "/agents/$name", params: { name } });
++          void navigate({ to: "/agents/$name", params: { name: resolvedAgentName } });
+         }}
+       />
+```
+</details>
 
 <details>
 <summary>🤖 Prompt for AI Agents</summary>
 
 ```
 Verify each finding against the current code and only fix it if needed.
 
-In `@internal/api/udsapi/agent_channels_test.go` around lines 72 - 116, Wrap the
-existing TestAgentCoordinatorConfigRouteReturnsResolvedPayload body in a subtest
-using t.Run("Should return resolved workspace coordinator payload", func(t
-*testing.T) { ... }) so the test follows the required t.Run("Should ...")
-pattern; locate the TestAgentCoordinatorConfigRouteReturnsResolvedPayload
-function and move its current contents into a t.Run call while keeping all setup
-(manager := activeAgentSessionManager, handlers := newTestHandlers,
-handlers.CoordinatorConfig = agentCoordinatorConfigResolverFunc, engine :=
-newTestRouter, performAgentKernelRequest, decodeJSONResponse and assertions)
-unchanged inside the subtest body.
+In `@web/src/routes/_app/agents`.$name.sessions.$id.tsx around lines 125 - 136,
+The onDeleteSuccess handler uses the possibly-stale route param name when
+navigating after deletion; change it to use the resolved agent name from the
+session (e.g., session.agent_name ?? name) so the redirect is canonical. Update
+the onDeleteSuccess in SessionPageContent's props to call navigate({ to:
+"/agents/$name", params: { name: session.agent_name ?? name } }) (or compute a
+resolvedAgentName variable) to ensure consistent post-delete routing.
 ```
 
 </details>
@@ -45,4 +60,9 @@ unchanged inside the subtest body.
 ## Triage
 
 - Decision: `VALID`
-- Notes: `TestAgentCoordinatorConfigRouteReturnsResolvedPayload` contains one direct top-level test body while this repo requires each case to run under a `t.Run("Should ...")` subtest. Fix by moving the existing setup, request, and assertions into a named `Should return resolved workspace coordinator payload` subtest.
+- Notes:
+  - The route passes `session.agent_name ?? name` to the page content but still navigates after delete with the raw URL param `name`.
+  - If the URL param is stale or manually edited, delete success redirects to a non-canonical agent route even though the resolved session payload contains the correct agent name.
+  - Fix by computing a single resolved agent name from the session and using it for both rendering and post-delete navigation. Add a route test that exercises delete success with a mismatched URL param.
+  - Resolution: reused `resolvedAgentName` for both content and delete-success navigation, with a regression test covering a mismatched route param.
+  - Verification: targeted Vitest passed; `make verify` passed.

.compozy/tasks/qa-rounds/reviews-001/issue_003.md

@@ -1,93 +1,26 @@
 ---
 status: resolved
-file: internal/automation/dispatch.go
-line: 60
+file: web/src/routes/_app/agents.$name.tsx
+line: 18
+severity: nitpick
 author: coderabbitai[bot]
-provider_ref: thread:PRRT_kwDOR5y4QM59r7vH,comment:PRRC_kwDOR5y4QM67Z0NB
+provider_ref: review:4177569108,nitpick_hash:803039727d80
+review_hash: 803039727d80
+source_review_id: "4177569108"
+source_review_submitted_at: "2026-04-26T22:35:58Z"
 ---
 
-# Issue 003: _🛠️ Refactor suggestion_ | _🟠 Major_
+# Issue 003: This still fetches agent detail data on nested session routes.
 ## Review Comment
 
-_🛠️ Refactor suggestion_ | _🟠 Major_
-
-**Make session stop timeout configurable, not hardcoded.**
-
-The new 10s value is operational policy in a core runtime path; it should be injected via dispatcher options (or TOML-backed config), not fixed in code.
-
-
-<details>
-<summary>Proposed refactor</summary>
-
-```diff
--const dispatcherSessionStopTimeout = 10 * time.Second
-+const defaultDispatcherSessionStopTimeout = 10 * time.Second
-
- type Dispatcher struct {
-   sessions SessionCreator
-   runs     RunStore
-   tasks    TaskService
-+  sessionStopTimeout time.Duration
-   ...
- }
-
- func NewDispatcher(sessions SessionCreator, runs RunStore, opts ...DispatcherOption) (*Dispatcher, error) {
-   ...
-   dispatcher := &Dispatcher{
-     sessions:      sessions,
-     runs:          runs,
-     logger:        slog.Default(),
-     now:           func() time.Time { return time.Now().UTC() },
-     sleep:         sleepWithContext,
-     maxConcurrent: DefaultMaxConcurrentJobs,
-+    sessionStopTimeout: defaultDispatcherSessionStopTimeout,
-   }
-   ...
- }
-
-+func WithDispatcherSessionStopTimeout(timeout time.Duration) DispatcherOption {
-+  return func(dispatcher *Dispatcher) {
-+    if timeout > 0 {
-+      dispatcher.sessionStopTimeout = timeout
-+    }
-+  }
-+}
-
- func (d *Dispatcher) stopAutomationSession(ctx context.Context, sessionID string, status RunStatus, runErr error) error {
-   ...
--  stopCtx, cancel := context.WithTimeout(context.WithoutCancel(ctx), dispatcherSessionStopTimeout)
-+  stopCtx, cancel := context.WithTimeout(context.WithoutCancel(ctx), d.sessionStopTimeout)
-   defer cancel()
-   ...
- }
-```
-</details>
-As per coding guidelines, "Never hardcode configuration in Go — use TOML config or functional options".
-
-<details>
-<summary>🤖 Prompt for AI Agents</summary>
-
-```
-Verify each finding against the current code and only fix it if needed.
-
-In `@internal/automation/dispatch.go` around lines 56 - 60, The hardcoded constant
-dispatcherSessionStopTimeout should be made configurable via dispatcher options
-rather than fixed in code: remove or replace the package-level const
-dispatcherSessionStopTimeout and add a field (e.g., SessionStopTimeout
-time.Duration) to the dispatcher options/config struct used by NewDispatcher (or
-Dispatcher) and its option helpers; wire that value into the shutdown logic that
-currently references dispatcherSessionStopTimeout and provide a sensible default
-(10*time.Second) when the option is not set so existing behavior remains
-unchanged.
-```
-
-</details>
-
-<!-- fingerprinting:phantom:poseidon:hawk -->
-
-<!-- This is an auto-generated comment by CodeRabbit -->
+`useAgentDetailPage(name)` runs before the child-route early return, so `/agents/$name/sessions/$id` still pays for the agent + sessions queries even though this component renders only `<Outlet />`. The clean fix is to move the detail shell into an index child route and let this parent route stay as a layout route.
 
 ## Triage
 
 - Decision: `VALID`
-- Notes: `dispatcherSessionStopTimeout` is a package-level operational timeout fixed at 10 seconds. The dispatcher already uses functional options for policy injection, so this should be configurable with a default. Fix by adding a `sessionStopTimeout` field, a `defaultDispatcherSessionStopTimeout`, and `WithDispatcherSessionStopTimeout`, then using the field in `stopAutomationSession`. A focused constructor test in `internal/automation/dispatch_test.go` is needed even though that file is outside the batch list because it validates the new option.
+- Notes:
+  - `AgentDetailPage` currently calls `useAgentDetailPage(name)` before checking `useChildMatches()`, so nested session routes still start agent detail and session list queries even though the parent renders only `<Outlet />`.
+  - The root cause is that the detail data hook lives in the layout component instead of in the detail-only rendering branch.
+  - Fix by splitting the detail shell into a child component that is rendered only when there is no child match, preserving hook rules while preventing nested routes from paying for unused data. Add a route test proving child routes render the outlet without invoking `useAgentDetailPage`.
+  - Resolution: split the detail shell into `AgentDetailContent`, leaving nested child routes to render `<Outlet />` without invoking the detail hook.
+  - Verification: targeted Vitest passed; `make verify` passed.