refactor(perf): heavy screens + analytics

[Wellington01]

No description provided.

[kody-ai]

Code Review Completed! 🔥

The code review was successfully completed based on your current configurations.

Kody Guide: Usage and Configuration

Interacting with Kody

• Request a Review: Ask Kody to review your PR manually by adding a comment with the `@kody start-review` command at the root of your PR.

• Provide Feedback: Help Kody learn and improve by reacting to its comments with a 👍 for helpful suggestions or a 👎 if improvements are needed.

Current Kody Configuration

Review Options

The following review options are enabled or disabled:

Options	Enabled
Bug	✅
Performance	✅
Security	✅
Business Logic	✅

Kody Code Review — 6 suggested fixes.
Paste the prompt below to your agent and all review fixed at once!

🛠️ Open Agent Prompt

A code review identified the following issues in this pull request.
Each section describes what was found and includes a reference implementation where available.

Files involved:
- apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:156
- apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:156
- apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:122
- libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
- libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
- libs/sandbox/infrastructure/providers/local-sandbox.service.ts:814

---

### [1/6] apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:156
Issue identified during code review:
In `page.client.tsx`, the mine view pins the author filter to the sentinel `me`, but the backend applies author filtering in memory after fetching automation executions while pagination totals are computed from the unfiltered execution count. This produces incorrect `totalItems`, `totalPages`, and `hasNextPage` values, can load empty pages, and forces repeated database round-trips to collect the user's PRs, so push the `me` filter into the execution repository query so counts and pages are computed against the same filtered result set.

---

### [2/6] apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:156
Issue identified during code review:
In `page.client.tsx`, the mine view pins the author to `me`, but this is filtered post-query after the backend has already fetched and paginated a batch of team-wide executions. Because contributors now default to this view, a contributor in a large team can receive pages containing none or only a few of their PRs and require multiple backend round-trips before the infinite list reaches 30 visible items, so push the `me` filter down to the Postgres/Mongo query so pagination happens after author scoping, or add a dedicated backend endpoint that scopes executions to the current user before applying LIMIT/OFFSET.
Reference implementation (from code review):

// apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:156
// "Minha fila" pins the list to my own PRs. `me` is a backend sentinel the
    // enriched use-case resolves to the logged-in user's git identity, so it
    // overrides any free-text author search while this view is active.
    const effectiveAuthor = isMineView
        ? "me"
        : debouncedAuthor.trim() || undefined;
    // TODO: author='me' is filtered post-query in the enriched use-case;
    // consider pushing this filter to the execution query so pagination is
    // scoped before LIMIT/OFFSET and mine-view pages are not under-filled.

---

### [3/6] apps/web/src/app/(app)/pull-requests/_components/page.client.tsx:122
Issue identified during code review:
Move role-to-permission decisions into `role-policies.ts` instead of doing an inline role comparison in a component, and use a policy/permission helper to determine the default scope. Run `yarn permissions:matrix` to regenerate the docs matrix if any permissions change.

---

### [4/6] libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
Issue identified during code review:
In 'mine' scope, the needsAttention facet in `get-pull-requests-facets.use-case.ts` is scoped to the caller's email without a falsy-email guard, so a missing email causes it to count the whole team's open actionable PRs instead of zero. This makes the 'mine' view disagree with the mine facet, which already returns 0 when email is absent, so return 0 for needsAttention in 'mine' scope when email is undefined.
Reference implementation (from code review):

// libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
const email = this.request.user?.email;
            // In the "mine" scope the actionable card counts only my open PRs
            // with an unaddressed suggestion — same predicate, plus the author.
            // countDeliveredPullRequests ANDs authorEmail with the other flags,
            // so this reuses the existing query (no new scan).
            const needsAttentionAuthor =
                query.scope === 'mine' ? email : undefined;

            const [reviewedKeys, needsAttention, mine, awaitingKeys] =
                await Promise.all([
                    this.automationExecutionService.getDistinctReviewedPullRequestKeys(
                        { organizationAndTeamData, repositoryIds },
                    ),
                    // Needs attention = open PRs that still carry an unaddressed
                    // delivered suggestion (any severity, implementationStatus ≠
                    // implemented). Actionable "someone needs to look at this",
                    // not "ever delivered a crit/high" (which counted merged/
                    // resolved PRs too). Scoped to the caller in the "mine" view.
                    needsAttentionAuthor || query.scope !== 'mine'
                        ? this.pullRequestsService.countDeliveredPullRequests(
                              organizationId,
                              repositoryIds,
                              {
                                  unresolvedOnly: true,
                                  openOnly: true,
                                  authorEmail: needsAttentionAuthor,
                              },
                          )
                        : Promise.resolve(0),
                    email
                        ? this.pullRequestsService.countDeliveredPullRequests(
                              organizationId,
                              repositoryIds,
                              { authorEmail: email },
                          )
                        : Promise.resolve(0),

---

### [5/6] libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
Issue identified during code review:
In `get-pull-requests-facets.use-case.ts`, `needsAttentionAuthor` is set to `request.user?.email` in 'mine' scope without a falsy guard, so a missing email causes `countDeliveredPullRequests` to omit the author filter and count the whole team's open unresolved PRs. The 'Needs my attention' card in the mine view then shows team-wide counts instead of zero, unlike the existing mine facet that already returns 0 when email is missing, so guard the call to return 0 in 'mine' scope when the caller has no email.
Reference implementation (from code review):

// libs/code-review/application/use-cases/dashboard/get-pull-requests-facets.use-case.ts:136
const email = this.request.user?.email;
            // In the "mine" scope the actionable card counts only my open PRs
            // with an unaddressed suggestion — same predicate, plus the author.
            // countDeliveredPullRequests ANDs authorEmail with the other flags,
            // so this reuses the existing query (no new scan).
            const needsAttentionAuthor =
                query.scope === 'mine' && email ? email : undefined;

            const [reviewedKeys, needsAttention, mine, awaitingKeys] =
                await Promise.all([
                    this.automationExecutionService.getDistinctReviewedPullRequestKeys(
                        { organizationAndTeamData, repositoryIds },
                    ),
                    // Needs attention = open PRs that still carry an unaddressed
                    // delivered suggestion (any severity, implementationStatus ≠
                    // implemented). Actionable "someone needs to look at this",
                    // not "ever delivered a crit/high" (which counted merged/
                    // resolved PRs too). Scoped to the caller in the "mine" view.
                    needsAttentionAuthor
                        ? this.pullRequestsService.countDeliveredPullRequests(
                              organizationId,
                              repositoryIds,
                              {
                                  unresolvedOnly: true,
                                  openOnly: true,
                                  authorEmail: needsAttentionAuthor,
                              },
                          )
                        : Promise.resolve(0),
                    email
                        ? this.pullRequestsService.countDeliveredPullRequests(
                              organizationId,
                              repositoryIds,
                              { authorEmail: email },
                          )
                        : Promise.resolve(0),

---

### [6/6] libs/sandbox/infrastructure/providers/local-sandbox.service.ts:814
Issue identified during code review:
The break condition of the `while (true)` loop in `local-sandbox.service.ts` uses strict equality (`parent === current`), which is discouraged because the loop only stops when the comparison is met exactly. This check is also redundant because `parentStrictlyInside` already becomes `false` once the parent reaches the repo root or escapes it, so remove the equality operand and rely on the relative-boundary guard instead.

---

Review each issue in context, use the reference implementations as guidance, and apply fixes that are consistent with the surrounding codebase.

Access your configuration settings here.

[kody-ai]

Found critical issues please review the requested changes

• Loop break condition uses strict equality (parent === current) in its termination check.

[kody-ai]

Code Review Completed! 🔥

The code review was successfully completed based on your current configurations.

Kody Guide: Usage and Configuration

Interacting with Kody

• Request a Review: Ask Kody to review your PR manually by adding a comment with the `@kody start-review` command at the root of your PR.

• Provide Feedback: Help Kody learn and improve by reacting to its comments with a 👍 for helpful suggestions or a 👎 if improvements are needed.

Current Kody Configuration

Review Options

The following review options are enabled or disabled:

Options	Enabled
Bug	✅
Performance	✅
Security	✅
Business Logic	✅

Kody Code Review — 1 suggested fix.
Paste the prompt below to your agent and all review fixed at once!

🛠️ Open Agent Prompt

A code review identified the following issues in this pull request.
Each section describes what was found and includes a reference implementation where available.

Files involved:
- apps/web/src/app/(app)/pull-requests/_components/pr-author-search.tsx:95

---

### [1/1] apps/web/src/app/(app)/pull-requests/_components/pr-author-search.tsx:95
Issue identified during code review:
The `onKeyDown` early return guards the `Escape` handler, so pressing Escape while the dropdown is open but empty does nothing, preventing users from keyboard-closing the autocomplete when the query returns no authors. Move the `Escape` branch before the `filtered.length === 0` guard, or remove `filtered.length === 0` from the guard and check it only for ArrowDown/ArrowUp/Enter.
Reference implementation (from code review):

// apps/web/src/app/(app)/pull-requests/_components/pr-author-search.tsx:95
onKeyDown={(event) => {
                    if (!open) return;
                    if (event.key === "Escape") {
                        setOpen(false);
                        return;
                    }
                    if (filtered.length === 0) return;
                    if (event.key === "ArrowDown") {

---

Review each issue in context, use the reference implementations as guidance, and apply fixes that are consistent with the surrounding codebase.

Access your configuration settings here.

[kody-ai]

Code Review Completed! 🔥

The code review was successfully completed based on your current configurations.

Kody Guide: Usage and Configuration

Interacting with Kody

• Request a Review: Ask Kody to review your PR manually by adding a comment with the `@kody start-review` command at the root of your PR.

• Provide Feedback: Help Kody learn and improve by reacting to its comments with a 👍 for helpful suggestions or a 👎 if improvements are needed.

Current Kody Configuration

Review Options

The following review options are enabled or disabled:

Options	Enabled
Bug	✅
Performance	✅
Security	✅
Business Logic	✅

Access your configuration settings here.

[kody-ai]

Code Review Completed! 🔥

The code review was successfully completed based on your current configurations.

Kody Guide: Usage and Configuration

Interacting with Kody

• Request a Review: Ask Kody to review your PR manually by adding a comment with the `@kody start-review` command at the root of your PR.

• Provide Feedback: Help Kody learn and improve by reacting to its comments with a 👍 for helpful suggestions or a 👎 if improvements are needed.

Current Kody Configuration

Review Options

The following review options are enabled or disabled:

Options	Enabled
Bug	✅
Performance	✅
Security	✅
Business Logic	✅

Access your configuration settings here.