feat(desktop): import cookies from Safari

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}

[juliusmarminge]

Stacked on #7261 (browser-import-linux-windows).

Adds Safari cookie import on macOS and a Full Disk Access step. System Settings opens through a dedicated main-process action with a fixed destination. After granting access, the wizard returns to configuration or retries the selected import, depending on where permission was needed.

The Safari reader validates the binary cookie structure, translates Apple-epoch timestamps, and preserves source-path context on parse errors. TCC denial is reported without attempting a bypass.

This layer retains the Windows-user-tested detection contribution #7323: modern/legacy Chromium cookie paths and Firefox’s active Windows lock probe. Access-denied errors are not treated as proof that Firefox is running. Windows Chromium remains intentionally unsupported.

Validation: binary parser fixtures and malformed-record boundaries, filesystem/lock fixtures, fixed System Settings destination, and permission-flow state transitions. Earlier desktop testing by the maintainer is preserved; no new live Safari or Windows GUI test was run in this audit.

Original implementation: Claude Code and external contribution #7323. Review fixes: GPT-5.6 Sol agents, coordinated through Codex.

Note

Add Safari cookie import with Full Disk Access detection and Linux Chromium key fix

• Adds a Safari source to BROWSER_IMPORT_SOURCES on macOS and parses Safari's Cookies.binarycookies format, converting Apple-epoch timestamps and extracting domain, path, security flags, and SameSite fields in SafariCookies.ts
• Introduces a needsFullDiskAccess unavailable reason; when an EPERM is hit opening the Safari jar, the import wizard in BrowserImportWizard.tsx shows a permission step that can open macOS System Settings and recheck access
• Wires an openSystemSettings IPC channel through the desktop bridge, preload, and ElectronShell to open the Privacy & Security Full Disk Access pane
• Fixes Linux Chromium key retrieval in ChromiumKeys.ts by adding the Chromium libsecret schema selector to secret-tool, so items with the same application attribute but a different schema are ignored
• Firefox profile discovery in Sources.ts now excludes declared profiles without a cookie database and falls back to directory scanning when no declared profile has one
• Risk: BrowserImportWizard callers must now provide a required openSystemSettings callback prop; the keychain-read audit log in BrowserImport.make is no longer emitted for non-macOS Chromium imports

^{Macroscope summarized d163c94.}

---

Note

Medium Risk
Touches cookie import, macOS TCC/Full Disk Access handling, and new IPC surface; parser mistakes could drop cookies silently, but the diff adds explicit bounds checks and tests for malformed jars.

Overview
Adds Safari as a macOS browser-import source by reading Cookies.binarycookies, with strict bounds checking on the binary format and mapping TCC EPERM denials to a new needsFullDiskAccess reason (distinct from ordinary EACCES failures).

Pre-flight listing probes Safari jar access via safariAccessDenied so the import wizard can land on a Full Disk Access step before configuration. That step can open macOS Privacy & Security → Full Disk Access through a new openSystemSettings("full-disk-access") path (fixed deep link in main process, pane id over IPC—not arbitrary URLs), wired from preload through localApi into IntegrationsSettings.

Browser import elsewhere: Chromium cookie DB resolution prefers Network/Cookies (127+) with legacy fallback; Firefox profile discovery skips empty/stale profiles.ini entries and rescans when needed; Windows treats stale Firefox parent.lock as not running and keeps Chromium forks off win32 (App-Bound Encryption). Keychain audit logging for Chromium imports is limited to darwin.

^{Reviewed by Cursor Bugbot for commit d163c94. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2ea8679c-247d-4066-a7c4-bc7ad8914eb6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ No successful main baseline artifact is available yet. This run establishes the initial measurement.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	—	13.3 KiB	—	15.1 KiB	✅
Codex	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Codex	Live turn WebSocket wire	—	6.4 KiB	—	7.8 KiB	✅
Codex	Live turn WebSocket decoded	—	55.6 KiB	—	66.4 KiB	✅
Codex	Live turn messages	—	10	—	21	✅
Claude	Total thread wire	—	13.1 KiB	—	15.1 KiB	✅
Claude	Thread snapshot wire	—	6.9 KiB	—	7.3 KiB	✅
Claude	Live turn WebSocket wire	—	6.2 KiB	—	7.8 KiB	✅
Claude	Live turn WebSocket decoded	—	55.5 KiB	—	66.4 KiB	✅
Claude	Live turn messages	—	8	—	21	✅

Baseline: unavailable · PR result: d163c94 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Not approved

Macroscope's review found this PR not approvable — Adds a substantial macOS Safari cookie-import capability with sensitive cookie-file access, a new Full Disk Access UI/IPC flow, and cross-platform browser-detection changes. The new test file also adds a file-level static-analysis suppression directive, so the change requires human review.

No code changes detected at d163c94. Prior analysis still applies.

^{You can add or adjust custom eligibility rules. Learn more.}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[macroscopeapp]

Effect service conventions review: one finding on the new Safari cookie error. Everything else (subpath namespace imports, Effect.fn boundaries, tagged-error definition, pass-through of the already structured error in Effect.try, no hidden runtimes) matches the conventions.

^{Posted via Macroscope — Effect Service Conventions}

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{Bugbot Autofix is ON, but a cloud agent failed to start.}

^{Reviewed by Cursor Bugbot for commit 6918217. Configure here.}