apps/desktop/src/main/ai/prompts/subtask-prompt-generator.ts (1)

332-335: ⚠️ Potential issue | 🟡 Minor

Missed rename: commit message prefix still uses auto-claude:.

The commit message template still references auto-claude: while the PR objective is to rename all references to aperant. This should be updated to maintain consistency with the rebranding.
,

🔧 Proposed fix

    `5. **Commit your changes:**\n` +
    `   \`\`\`bash\n` +
    `   git add .\n` +
-   `   git commit -m "auto-claude: ${subtask.id} - ${subtask.description.slice(0, 50)}"\n` +
+   `   git commit -m "aperant: ${subtask.id} - ${subtask.description.slice(0, 50)}"\n` +
    `   \`\`\`\n` +

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ai/prompts/subtask-prompt-generator.ts` around lines
332 - 335, The commit message template still uses the old prefix "auto-claude:";
update the string in the git commit line (the template that builds `git commit
-m "auto-claude: ${subtask.id} - ${subtask.description.slice(0, 50)}"`) to use
the new brand prefix "aperant:" so the commit message becomes `aperant:
${subtask.id} - ${subtask.description.slice(0, 50)}`.

apps/desktop/src/renderer/components/onboarding/GraphitiStep.tsx (1)

118-118: ⚠️ Potential issue | 🟠 Major

Incomplete branding update: auto_claude_memory defaults should be renamed to align with .aperant directory structure.

The file uses .aperant for the storage path (line 910) but retains auto_claude_memory for the database default (lines 118, 249) and placeholder (line 905). Update these to use a consistent branding name (e.g., aperant_memory or aperant). This pattern also appears in SecuritySettings.tsx:429, MemoryBackendSection.tsx:133, and other files—apply the same rename consistently across the codebase.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/onboarding/GraphitiStep.tsx` at line
118, Replace the outdated branding value "auto_claude_memory" with the new
consistent name (e.g., "aperant_memory") wherever it is used as a default or
placeholder in GraphitiStep.tsx (the database default entry and the placeholder
near the storage path usage), and make the same replacement in the other
referenced components (SecuritySettings.tsx, MemoryBackendSection.tsx) so
defaults, placeholders, and any related keys match the .aperant directory
branding; update any variable names/comments that reference auto_claude_memory
to the new identifier to keep the codebase consistent.

apps/desktop/src/main/ai/runners/__tests__/ideation.test.ts (1)

284-295: ⚠️ Potential issue | 🟡 Minor

Test name and assertions are out of sync (missing outputDir assertion).

The test claims to validate both projectDir and outputDir, but it only asserts project path content. Add an assertion for the .aperant/ideation output path to prevent regressions.

💡 Suggested test fix

   const streamArgs = mockStreamText.mock.calls[0][0];
   const systemPrompt = streamArgs.system as string;
   expect(systemPrompt).toContain('/my/project');
+  expect(systemPrompt).toContain('/my/project/.aperant/ideation');

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ai/runners/__tests__/ideation.test.ts` around lines 284
- 295, The test for runIdeation names both projectDir and outputDir but only
asserts the projectDir; update the test (using mockStreamText and inspecting
streamArgs.system as systemPrompt) to also assert that the outputDir is included
by adding an expectation that systemPrompt contains the output path
'/my/project/.aperant/ideation' (e.g.,
expect(systemPrompt).toContain('/my/project/.aperant/ideation')); leave the
existing projectDir assertion intact.

apps/desktop/src/renderer/components/AgentTools.tsx (1)

1186-1192: ⚠️ Potential issue | 🟡 Minor

Always-enabled row still uses legacy autoClaude translation keys.

If translations were renamed to aperant, this will surface stale text (or raw keys). Update this row to settings:mcp.servers.aperant.* for naming consistency with the rest of the migration.

Minimal key update

-<span className="text-sm font-medium">{t('settings:mcp.servers.autoClaude.name')}</span>
-<p className="text-xs text-muted-foreground">{t('settings:mcp.servers.autoClaude.description')} ({t('settings:mcp.alwaysEnabled')})</p>
+<span className="text-sm font-medium">{t('settings:mcp.servers.aperant.name')}</span>
+<p className="text-xs text-muted-foreground">{t('settings:mcp.servers.aperant.description')} ({t('settings:mcp.alwaysEnabled')})</p>

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/AgentTools.tsx` around lines 1186 -
1192, Update the stale translation keys in the AgentTools component: replace
t('settings:mcp.servers.autoClaude.name') and
t('settings:mcp.servers.autoClaude.description') with
t('settings:mcp.servers.aperant.name') and
t('settings:mcp.servers.aperant.description') respectively (the surrounding
layout including the always-enabled note t('settings:mcp.alwaysEnabled') stays
unchanged); these strings appear in the Auto-Claude/aperant row in
AgentTools.tsx where the span and paragraph render the name and description.

apps/desktop/src/main/ipc-handlers/sections/task-section.txt (1)

778-784: 🧹 Nitpick | 🔵 Trivial

Consider using getSpecsDir() helper for consistency.

Similar to the pattern at line 261, this code uses an inline fallback instead of the getSpecsDir() helper function used elsewhere in the file. Consolidating to use the helper would reduce code duplication and improve maintainability.

♻️ Proposed refactor to use helper function

-      const autoBuildDir = project.autoBuildPath || '.aperant';
-      const specDir = path.join(
-        project.path,
-        autoBuildDir,
-        'specs',
-        task.specId
-      );
+      const specsBaseDir = getSpecsDir(project.autoBuildPath);
+      const specDir = path.join(project.path, specsBaseDir, task.specId);

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ipc-handlers/sections/task-section.txt` around lines
778 - 784, Replace the inline autoBuildDir and specDir path assembly with the
existing getSpecsDir helper to match the pattern used at line 261: remove the
autoBuildDir and path.join usage that builds specDir from
project.autoBuildPath/project.path and task.specId, and instead call
getSpecsDir(...) (the same signature used elsewhere, e.g., getSpecsDir(project,
task.specId) or getSpecsDir(project) plus specId) to obtain the specs directory;
assign specDir from that helper and keep task.specId usage consistent.

apps/desktop/src/main/utils/worktree-cleanup.ts (1)

132-142: 🧹 Nitpick | 🔵 Trivial

Pre-existing: process.platform usage violates coding guidelines.

Line 134 uses process.platform !== 'win32' directly. Per coding guidelines, platform detection should use functions from apps/desktop/src/main/platform/ such as isWindows().

This is pre-existing code, but consider addressing it while touching this file.

♻️ Suggested refactor

+import { isWindows } from '../platform';
...
   // All retries exhausted - try shell rm -rf as fallback on Unix
   // Node's rm() can fail with ENOTEMPTY on macOS .app bundles
-  if (process.platform !== 'win32') {
+  if (!isWindows()) {
     try {

As per coding guidelines: "Never use process.platform directly. Import platform detection from apps/desktop/src/main/platform/ instead."

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/utils/worktree-cleanup.ts` around lines 132 - 142,
Replace the direct process.platform check with the centralized platform helper:
import and call isWindows() from the platform utilities and change the condition
to if (!isWindows()) around the block that logs and calls
execFileSync('/bin/rm', ['-rf', dirPath], ...); ensure you add the import for
isWindows at the top of the module and keep existing references to logPrefix and
dirPath unchanged so behavior remains the same.

apps/desktop/src/renderer/components/project-settings/SecuritySettings.tsx (1)

426-442: ⚠️ Potential issue | 🟠 Major

Replace hardcoded UI literals with react-i18next translation keys.

Lines 426, 438, and 441 contain hardcoded user-facing text in TSX. These must use react-i18next keys with entries in both en/settings.json and fr/settings.json translation files.

Suggested direction

+ import { useTranslation } from 'react-i18next';

 export function SecuritySettings(...) {
+  const { t } = useTranslation(['settings']);
   ...
-  <p className="text-xs text-muted-foreground">
-    Stored in ~/.aperant/memories/
-  </p>
+  <p className="text-xs text-muted-foreground">
+    {t('settings:security.memory.storedIn', { path: '~/.aperant/memories/' })}
+  </p>

-  <p className="text-xs text-muted-foreground">
-    Custom storage location. Default: ~/.aperant/memories/
-  </p>
+  <p className="text-xs text-muted-foreground">
+    {t('settings:security.memory.customLocationDefault', { path: '~/.aperant/memories/' })}
+  </p>

-  <Input placeholder="~/.aperant/memories" ... />
+  <Input placeholder={t('settings:security.memory.pathPlaceholder')} ... />

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/project-settings/SecuritySettings.tsx`
around lines 426 - 442, In SecuritySettings.tsx, replace the hardcoded
user-facing literals (the paragraph "Stored in ~/.aperant/memories/", the Label
"Database Path (Optional)", the paragraph "Custom storage location. Default:
~/.aperant/memories/", and the two Input placeholders "auto_claude_memory" and
"~/.aperant/memories") with react-i18next keys using useTranslation (e.g.,
t('settings.memoryStored'), t('settings.dbPathLabel'), t('settings.dbPathHelp'),
t('settings.placeholderMemory'), t('settings.placeholderDbPath')) and keep
wiring to envConfig.memoryDatabase and envConfig.memoryDbPath unchanged; then
add corresponding entries to en/settings.json and fr/settings.json with
appropriate translations. Ensure you import and call useTranslation at the top
of SecuritySettings and replace the literal strings with the t(...) calls.

apps/desktop/src/main/utils/git-isolation.ts (1)

166-169: 🧹 Nitpick | 🔵 Trivial

Minor: Comment still references "auto-claude branch"

The security comment on lines 166-169 mentions "another task's auto-claude branch" but the actual branch pattern is now aperant/. Consider updating for consistency.

📝 Suggested comment update

     // SECURITY: Use strict exact-match validation (not prefix matching) to prevent
-    // accidentally deleting a different task's auto-claude branch. When git rev-parse
+    // accidentally deleting a different task's aperant branch. When git rev-parse
     // returns an unexpected branch, we MUST fall back to the expected pattern rather
     // than risking deletion of the wrong branch. This is critical for data safety.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/utils/git-isolation.ts` around lines 166 - 169, Update
the security comment in git-isolation.ts so it references the current branch
naming scheme (`aperant/`) instead of "auto-claude branch"; specifically edit
the exact comment block that starts with "SECURITY: Use strict exact-match
validation..." to mention preventing deletion of another task's `aperant/`
branch and ensure the wording makes clear we must fall back to the expected
`aperant/` pattern when git rev-parse returns an unexpected branch.

apps/desktop/src/renderer/components/project-settings/MemoryBackendSection.tsx (1)

128-145: 🧹 Nitpick | 🔵 Trivial

Hardcoded strings violate i18n guidelines (pre-existing issue)

The user-facing text in this component uses hardcoded English strings rather than react-i18next translation keys. While the path updates from .auto-claude to .aperant are correct, these strings should use i18n for proper localization:

• Line 128: "Database Name"
• Line 130: "Name for the memory database (stored in ~/.aperant/memories/)"
• Line 140: "Database Path (Optional)"
• Line 142: "Custom storage location. Default: ~/.aperant/memories/"

This appears to be a pre-existing issue not introduced by this PR. Consider addressing in a follow-up.

As per coding guidelines: "All frontend user-facing text must use react-i18next translation keys. Hardcoded strings in JSX/TSX break localization for non-English users."

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In
`@apps/desktop/src/renderer/components/project-settings/MemoryBackendSection.tsx`
around lines 128 - 145, Replace the hardcoded English strings in
MemoryBackendSection.tsx with react-i18next translation keys: import and call
useTranslation (e.g., const { t } = useTranslation()) at the top of the
component and replace "Database Name", "Name for the memory database (stored in
~/.aperant/memories/)", "Database Path (Optional)", and "Custom storage
location. Default: ~/.aperant/memories/" with t('memory.databaseName'),
t('memory.databaseDescription'), t('memory.databasePathOptional'), and
t('memory.databasePathDescription') (and also translate placeholders like
"auto_claude_memory" and "~/.aperant/memories"); ensure the new keys are added
to the locale resources in the i18n files in a follow-up change.

apps/desktop/src/renderer/lib/mocks/project-mock.ts (1)

9-16: ⚠️ Potential issue | 🟡 Minor

Update mock autoBuildPath to .aperant for consistency.

autoBuildPath still points to .../auto-claude (Line 15), which conflicts with this PR’s canonical folder rename and can skew renderer/mock behavior.

Suggested fix

-      autoBuildPath: `${projectPath}/auto-claude`,
+      autoBuildPath: `${projectPath}/.aperant`,

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/lib/mocks/project-mock.ts` around lines 9 - 16, The
mock addProject function returns an object whose autoBuildPath is hardcoded to
`${projectPath}/auto-claude`; update this to use the new canonical folder name
`.aperant` by changing the autoBuildPath generation in addProject to
`${projectPath}/.aperant` (keep other fields like id, name, path and settings
using DEFAULT_PROJECT_SETTINGS unchanged) so mock behavior matches the PR
rename.

apps/desktop/src/main/ai/config/agent-configs.ts (1)

477-486: ⚠️ Potential issue | 🟠 Major

Keep the legacy auto-claude alias in MCP_SERVER_NAME_MAP.

mapMcpServerName() resolves user-provided MCP override strings. Dropping the old name here turns the rename into a silent breaking change for any existing config that still says auto-claude.

🛠️ Proposed fix

 const MCP_SERVER_NAME_MAP: Record<string, string> = {
   context7: 'context7',
   'graphiti-memory': 'memory',
   graphiti: 'memory',
   memory: 'memory',
   linear: 'linear',
   electron: 'electron',
   puppeteer: 'puppeteer',
+  'auto-claude': 'aperant',
   'aperant': 'aperant',
 };

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ai/config/agent-configs.ts` around lines 477 - 486, The
MCP_SERVER_NAME_MAP is missing the legacy "auto-claude" alias which
mapMcpServerName() still needs to resolve older configs; add an entry for
"auto-claude" mapping to the current server key (e.g. map "auto-claude" ->
"context7") inside MCP_SERVER_NAME_MAP so older user-provided MCP override
strings continue to work and are not silently broken.

apps/desktop/src/renderer/components/AddProjectModal.tsx (1)

121-121: ⚠️ Potential issue | 🟡 Minor

Hardcoded error string violates i18n requirements.

The fallback string 'Failed to create project folder' should use a translation key for localization consistency. As per coding guidelines, all frontend user-facing text must use react-i18next translation keys.

🌐 Proposed fix

-        setError(result.error || 'Failed to create project folder');
+        setError(result.error || t('addProject.failedToCreate'));

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/AddProjectModal.tsx` at line 121, The
hardcoded fallback in setError(result.error || 'Failed to create project
folder') violates i18n rules; replace the literal with a translation lookup (use
the component's i18n hook: e.g. t('projects.createFailed') or an existing key)
and ensure useTranslation is imported/used in AddProjectModal.tsx if not already
present; keep result.error when present, but call setError(result.error ||
t('projects.createFailed')) so the UI uses a translation key for the default
message.

apps/desktop/src/renderer/components/WelcomeScreen.tsx (1)

104-108: ⚠️ Potential issue | 🟡 Minor

Hardcoded string "Initialized" breaks localization.

Line 106 contains a hardcoded English string which violates the coding guideline requiring all frontend user-facing text to use react-i18next translation keys. The rest of this component correctly uses the t() function.

🌐 Proposed fix to use translation key

 {project.aperantPath && (
   <span className="text-[10px] px-1.5 py-0.5 rounded-full bg-success/20 text-success shrink-0">
-    Initialized
+    {t('welcome:recentProjects.initialized')}
   </span>
 )}

Add the key to both translation files:

en/welcome.json:

"recentProjects": {
  ...
  "initialized": "Initialized"
}

fr/welcome.json:

"recentProjects": {
  ...
  "initialized": "Initialisé"
}

As per coding guidelines: "All frontend user-facing text must use react-i18next translation keys. Hardcoded strings in JSX/TSX break localization for non-English users."

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/WelcomeScreen.tsx` around lines 104 -
108, Replace the hardcoded "Initialized" string in WelcomeScreen.tsx with a
react-i18next lookup (use the t function) — specifically change the span that
renders when project.aperantPath to call t('recentProjects.initialized') instead
of the literal; then add the key "recentProjects.initialized": "Initialized" to
the English translation file and "recentProjects.initialized": "Initialisé" to
the French translation file so both locales are covered.

apps/desktop/src/main/utils/git-isolation.ts (1)

50-52: ⚠️ Potential issue | 🟡 Minor

Inconsistent naming: "Auto-Claude" reference remains in comment.

The comment on line 51 still references "Auto-Claude manages commits" while this PR renames all references to "Aperant". This should be updated for consistency.

📝 Proposed fix

  * Also sets HUSKY=0 to disable the user's pre-commit hooks when
- * Auto-Claude manages commits, preventing double-hook execution
+ * Aperant manages commits, preventing double-hook execution
  * and potential conflicts.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/utils/git-isolation.ts` around lines 50 - 52, Update
the inline comment that currently reads "Auto-Claude manages commits" to use the
new product name "Aperant" for consistency; specifically change the sentence
that mentions HUSKY=0 disabling the user's pre-commit hooks so it reads that
HUSKY=0 is set when "Aperant manages commits", leaving the surrounding text
(including the HUSKY=0 reference) unchanged.

apps/desktop/src/main/ipc-handlers/changelog-handlers.ts.bk (1)

1-264: ⚠️ Potential issue | 🟠 Major

Remove backup file from the repository.

This .bk file appears to be an accidentally committed backup and should be deleted. Backup files should be excluded via .gitignore to prevent accidental commits.

#!/bin/bash
# Check if there are other backup files accidentally committed
fd -e bk -e bak -e orig -e backup

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ipc-handlers/changelog-handlers.ts.bk` around lines 1 -
264, Delete the accidentally committed backup file that contains the duplicate
registerChangelogHandlers implementation (the .bk file); remove it from the
repository (git rm) and update .gitignore to exclude common backup extensions
(bk, bak, orig, backup) so future backups aren’t committed, and ensure the
canonical changelog handler implementation (function registerChangelogHandlers)
remains in the repo.

apps/desktop/src/renderer/components/GitHubSetupModal.tsx (2)

747-748: ⚠️ Potential issue | 🟡 Minor

Use translation key instead of hardcoded text.

This user-facing description text is hardcoded in English, breaking localization for non-English users. As per coding guidelines, all frontend user-facing text in renderer components must use react-i18next translation keys. Add a key to both en/*.json and fr/*.json translation files.

🌐 Suggested fix using translation key

 <DialogDescription>
-  Choose which branch Aperant should use as the base for creating task branches.
+  {t('githubSetup.branchSelectionDescription')}
 </DialogDescription>

Then add to translation files:

// en/dialogs.json
"githubSetup": {
  "branchSelectionDescription": "Choose which branch Aperant should use as the base for creating task branches."
}

// fr/dialogs.json
"githubSetup": {
  "branchSelectionDescription": "Choisissez la branche qu'Aperant doit utiliser comme base pour créer les branches de tâches."
}

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/GitHubSetupModal.tsx` around lines 747 -
748, Replace the hardcoded English description inside GitHubSetupModal (the
DialogDescription element) with a react-i18next translation key; use the
component's i18n instance (e.g., t('githubSetup.branchSelectionDescription'))
where DialogDescription currently contains the literal string. Add the
corresponding key "githubSetup.branchSelectionDescription" with the provided
English and French values to the renderer translation files (en/dialogs.json and
fr/dialogs.json). Ensure the component imports and uses the useTranslation hook
if not already present and update any tests or snapshot text if necessary.

---

798-804: ⚠️ Potential issue | 🟡 Minor

Use translation key with interpolation for help text.

This user-facing help text is hardcoded in English. The text should use a translation key that supports interpolation for the branch example and selected branch name.

🌐 Suggested fix using translation key with interpolation

 <p className="text-xs text-muted-foreground">
-  All tasks will be created from branches like{' '}
-  <code className="px-1 bg-muted rounded">aperant/task-name</code>
-  {selectedBranch && (
-    <> based on <code className="px-1 bg-muted rounded">{selectedBranch}</code></>
-  )}
+  {t('githubSetup.branchPatternHelp', {
+    pattern: 'aperant/task-name',
+    baseBranch: selectedBranch
+  })}
 </p>

Then add to translation files with proper handling of code elements:

// en/dialogs.json
"githubSetup": {
  "branchPatternHelp": "All tasks will be created from branches like <code>{{pattern}}</code>{{baseBranch, ' based on <code>{{baseBranch}}</code>'}}"
}

Note: You may need to use Trans component from react-i18next for proper HTML rendering of <code> elements.

As per coding guidelines, all frontend user-facing text must use react-i18next translation keys. Add keys to both en/*.json and fr/*.json translation files.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/GitHubSetupModal.tsx` around lines 798 -
804, The help text in GitHubSetupModal (the paragraph containing the branch
pattern and selectedBranch) is hardcoded; replace it with a react-i18next
translation key (e.g., "githubSetup.branchPatternHelp") that supports
interpolation and HTML/code elements using the Trans component, passing pattern
and selectedBranch as interpolation variables (use selectedBranch only when
present); add the corresponding key to en and fr translation JSON files
(ensuring markup for <code> is preserved or handled via Trans) and update the
component to render Trans with values={{ pattern: 'aperant/task-name',
baseBranch: selectedBranch }} so the branch example and optional "based on"
segment are localized and properly interpolated.

apps/desktop/src/main/utils/spec-number-lock.ts (1)

150-167: ⚠️ Potential issue | 🟠 Major

Replace hardcoded .aperant with variable to match dynamic path handling in main specs scan.

Line 157 hardcodes .aperant while line 153 respects the specsBase variable. If aperantPath ever differs from the default (via initialization or updates), worktree specs scanning will fail. Use the same variable reference for consistency.

💡 Proposed fix

     const specsBase = aperantPath || '.aperant';

     // 1. Scan main project specs
     const mainSpecsDir = path.join(this.projectDir, specsBase, 'specs');
     maxNumber = Math.max(maxNumber, this.scanSpecsDir(mainSpecsDir));

     // 2. Scan all worktree specs
-    const worktreesDir = path.join(this.projectDir, '.aperant', 'worktrees', 'tasks');
+    const worktreesDir = path.join(this.projectDir, specsBase, 'worktrees', 'tasks');
     if (existsSync(worktreesDir)) {

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/utils/spec-number-lock.ts` around lines 150 - 167, The
worktree scan uses a hardcoded '.aperant' path while the initial main specs scan
uses the dynamic specsBase variable; update the worktree path construction to
use specsBase instead of the literal '.aperant' so worktreeDir and
worktreeSpecsDir are built consistently (replace occurrences that join
'.aperant' with specsBase in the code around scanSpecsDir, worktreesDir and
worktreeSpecsDir where variables like specsBase, worktreesDir, scanSpecsDir and
worktree.name are used).

apps/desktop/src/main/ipc-handlers/task/crud-handlers.ts (1)

175-195: ⚠️ Potential issue | 🟠 Major

Use the global spec-number allocator here too.

This scan only inspects the main specs directory. The store already supports worktree-only specs, so a higher-numbered task that only exists in a worktree can be missed and this handler can reuse its number. Reuse the same global numbering strategy as ideation/task-converter.ts here.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ipc-handlers/task/crud-handlers.ts` around lines 175 -
195, The handler currently computes specNumber by scanning only the main
specsDir (using getSpecsDir and the existingDirs/existingNumbers logic), which
can clash with worktree-only specs; replace that local directory-only allocation
with the global spec-number allocator used in ideation/task-converter.ts: import
and call the allocator (the function exported by ideation/task-converter.ts that
returns the next global spec number) to set specNumber instead of the
existingDirs -> existingNumbers logic, and remove or fall back from the local
scan (the block that computes existingDirs, existingNumbers, and sets
specNumber) so the handler uses the same global numbering strategy as the rest
of the app.

apps/desktop/src/main/ipc-handlers/insights-handlers.ts (1)

81-84: ⚠️ Potential issue | 🟡 Minor

Remove console.log statements from production code.

Per coding guidelines, console.log output is invisible in bundled Electron apps. These debug statements should be removed or replaced with proper observability tooling.

🛠️ Suggested fix

-      console.log("[Insights Handler] Using model config:", {
-        model: configWithSettings.model,
-        thinkingLevel: configWithSettings.thinkingLevel,
-      });

As per coding guidelines: "Do not use console.log in production code. console.log output is invisible in bundled Electron apps."

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/main/ipc-handlers/insights-handlers.ts` around lines 81 -
84, Remove the console.log call in the Insights handler that prints
configWithSettings.model and configWithSettings.thinkingLevel; instead use the
application's structured logging facility (e.g., replace with
processLogger.debug/processLogger.info or the project's main logger) or drop the
debug entirely so no console logging remains in production. Update the code
referencing configWithSettings.model and configWithSettings.thinkingLevel in the
same block (the console.log shown) to call the chosen logger with a descriptive
message and include the values as structured fields rather than using
console.log.

apps/desktop/src/renderer/components/project-settings/GeneralSettings.tsx (2)

20-25: 🛠️ Refactor suggestion | 🟠 Major

Use @shared/* in this touched import block.

These ../../../shared/* imports make the component brittle to moves/renames. Please switch this block to the repo’s shared path aliases while you’re touching it.

♻️ Suggested change

-import { AVAILABLE_MODELS } from '../../../shared/constants';
+import { AVAILABLE_MODELS } from '@shared/constants';
 import type {
   Project,
   ProjectSettings as ProjectSettingsType,
   AperantVersionInfo
-} from '../../../shared/types';
+} from '@shared/types';

As per coding guidelines: apps/desktop/src/**/*.{ts,tsx}: Use path aliases defined in tsconfig.json for imports: @/* (renderer), @shared/* (shared), @preload/* (preload), @features/*, @components/*, @hooks/*, @lib/*.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/project-settings/GeneralSettings.tsx`
around lines 20 - 25, The import block in GeneralSettings.tsx uses relative
paths ('../../../shared/constants' and '../../../shared/types') which are
brittle; replace those with the shared path alias imports using `@shared` — e.g.
import { AVAILABLE_MODELS } from '@shared/constants' and import the types
Project, ProjectSettings (as ProjectSettingsType), AperantVersionInfo from
'@shared/types' — update the import statements referencing AVAILABLE_MODELS,
Project, ProjectSettingsType, and AperantVersionInfo accordingly so the
component uses the tsconfig path alias `@shared/`*.

---

50-76: ⚠️ Potential issue | 🟠 Major

Localize the new Aperant copy.

This block adds new English text directly in JSX even though the component already uses useTranslation. Move the new heading/help/button text into settings keys and add matching entries in both locales.

As per coding guidelines: apps/desktop/src/renderer/**/*.{tsx,jsx}: All frontend user-facing text must use react-i18next translation keys. Hardcoded strings in JSX/TSX break localization for non-English users. Add keys to both en/*.json and fr/*.json translation files.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/renderer/components/project-settings/GeneralSettings.tsx`
around lines 50 - 76, The new hardcoded strings in GeneralSettings.tsx (heading
"Aperant Integration", "Not Initialized", helper copy "Initialize Aperant to
enable task creation and agent workflows.", and button labels "Initializing..."
/ "Initialize Aperant") must be replaced with react-i18next keys from the
settings namespace; update the JSX to use t('settings.aperant.title'),
t('settings.aperant.statusNotInitialized'), t('settings.aperant.help'), and
t('settings.aperant.initializeButton' / 'settings.aperant.initializing') (or
similar) and keep using existing handlers like handleInitialize and isUpdating;
then add the corresponding entries under the settings section in both locales
(en JSON and fr JSON) so both languages have these new keys.

apps/desktop/src/shared/constants/config.ts (1)

98-119: ⚠️ Potential issue | 🟡 Minor

Remove the unused MEMORY_STATE constant or update its path to match the documented .aperant/ structure.

APERANT_PATHS.MEMORY_STATE appears to be dead code. A codebase search shows it's never referenced anywhere, while memory state is actually managed via getMemoriesDir() in config-paths.ts, which correctly handles migration from .auto-claude/memories to .aperant/memories. Either remove the constant or, if retained for backward compatibility, update it to '.aperant/.memory_state.json' and add a comment explaining its purpose.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@apps/desktop/src/shared/constants/config.ts` around lines 98 - 119,
APERANT_PATHS currently defines MEMORY_STATE as '.memory_state.json' but it's
unused and inconsistent with the documented .aperant/ layout; either remove
APERANT_PATHS.MEMORY_STATE or update it to '.aperant/.memory_state.json' and add
a short comment explaining it's kept for backward compatibility and that real
memory handling is done via getMemoriesDir() in config-paths.ts (which migrates
from .auto-claude/memories to .aperant/memories); modify the APERANT_PATHS
object accordingly and run a code search to ensure no references break.