Two fixes:
1. grok.com missing from manifest
Grok moved from grok.x.ai to grok.com. The extension wasn't activating
on grok.com at all (no content script injected). Added to host_permissions
and content_scripts in both manifests.
2. Fetch hook bypassed on React/Next.js sites
content.js was injected async (after awaiting the document-scanner),
so by the time it replaced window.fetch, React/Next.js had already stored
a reference to the original. All fetch calls from framework code bypassed
the hook entirely.
Fix: load early-hook.js as a "world": "MAIN" content script at
document_start. This runs synchronously before any page JavaScript,
captures the real fetch in window.__ssOriginalFetch, and replaces
window.fetch with a lightweight wrapper. When content.js eventually
loads, it sets window.__ssInterceptFetch (the real substitution logic)
and window.__ssReady = true, activating the wrapper. React's stored
fetch reference now routes through the interceptor.
Falls back to direct window.fetch replacement if early-hook.js somehow
didn't run (e.g. older browser without world: MAIN support).
https://claude.ai/code/session_01CwcZK8nqL8pyBH9AxDs9qo
- document-scanner.js: Add missing await on _inflateSync() call — FlateDecode
streams were not being decompressed because the async function was called
without await, causing decompressed to hold a Promise instead of data
- content.js, storage.js, options.html, options.js: Gate proper noun / capital
letter detection behind a new detectProperNouns setting (default off); was
previously always-on causing noisy false-positive warnings on AI thinking
output and common capitalized words
- Bump version 0.9.23 → 0.9.25
https://claude.ai/code/session_01ReZUeR1nrYzJeX7fTqwXMw
- Inject document-scanner.js into page world via injector.js (as module,
sets globalThis.DocumentScanner)
- Add FormData interception to fetch hook: scans File/Blob entries through
DocumentScanner.processUpload(), also substitutes string fields
- Add document-scanner.js to web_accessible_resources in both manifests
- Bump version to 0.9.23 in package.json, manifest.json, manifest.firefox.json
https://claude.ai/code/session_01NNBEPuXMFGWezJb1f958nL
Going back to a known-good baseline. This version had:
- Working reveal mode with CSS Highlight API
- Working substitution (fetch + XHR hooks)
- Smart patterns (names, emails, phones, usernames)
- Encryption/sync (password, TOTP, WebAuthn)
- Multiple identity profiles
- Activity log
- Secret scanner
- Auto-detect PII warnings
- Pre-send PII detection
Kept current manifests (UUID, data_collection_permissions, version).
No renames applied — uses original naming (secretScanning, PPI, etc).
Will re-apply renames and new features from this working base.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
The data-ss-config attribute on the script tag was being removed
(via script.onload) before content.js could read it. Changed approach:
inject config as a separate <script type="application/json" id="ss-config-data">
element that persists in the DOM until content.js reads and removes it.
This eliminates the race condition between script execution and onload
removal. Bump 0.9.20.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
Restored injector.js from v2.0.14 (commit 9a11896) which:
- Requests decrypted config from background via get:decrypted-config
message instead of passing empty arrays when data is encrypted
- Handles the identity.profiles merge correctly for background responses
- Passes ss_settings directly (not checking _ssLocalEncrypted which
caused settings loss)
Added missing get:decrypted-config message handler to service-worker.js
which returns decrypted mappings, identity, and settings via Storage
module.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
Reverted content.js to commit 9a11896 (the last version where reveal
mode worked) and re-applied only the renames:
- PPI → PII
- secretScanning → autoRedact
- SECRET_PATTERNS → REDACT_PATTERNS
- scanAndRedactSecrets → runAutoRedact
- category: 'secret' → category: 'redact'
- Added customRedactPatterns support to runAutoRedact
This restores:
- Individual word storage in sessionSubstitutions (needed for reveal)
- fetch(Request) handling (not just fetch(url, options))
- Blob/ArrayBuffer/URLSearchParams body conversion
- isInNonChatArea scoping (narrowed version)
- Proper noun detection gated by settings.detectProperNouns
- Mapping values added to configured skip set
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
The class-based selectors ([class*="header"], [class*="Header"],
[class*="nav-"], etc.) were matching Claude.ai's chat content area
elements, preventing reveal from running on any response text.
Narrowed to only structural elements: nav, aside, [role="navigation"],
[role="complementary"], [data-sidebar]. Removed HEADER/FOOTER from
SKIP_REVEAL_TAGS since sites use these tags inside chat layouts.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
The smart engine records individual name parts ("Ademo"→"John",
"Demo"→"Smith") AND the combined form ("Ademo Demo"→"John Smith")
in sessionSubstitutions. The catch-all in buildRevealPairs was
adding all of them, causing partial replacements that corrupted
the DOM and made the cache oscillate between 4 and 0 pairs.
Fix: skip session entries whose key is a substring of a longer
entry (e.g. "ademo" is part of "ademo demo"). Only the combined
form gets added as a reveal pair.
Also: remove debug logging, improve cache with size tracking.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
Root cause found: commit 9a11896 (v2.0.14) added isInNonChatArea()
and expanded SKIP_REVEAL_TAGS with NAV/ASIDE/HEADER/FOOTER, but that
commit was on a branch that was never merged into main. When we started
our work from main, these fixes were missing.
Restored:
- isInNonChatArea() function — prevents reveal from touching sidebars,
navigation, headers, footers, and other non-chat UI
- SKIP_REVEAL_TAGS expanded with NAV, ASIDE, HEADER, FOOTER
- isInNonChatArea checks added to revealInElement, unrevealInElement,
and highlightMatches (both element-level and walker-level)
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
Two bugs found:
1. Individual name parts ("Ademo"→"John", "Demo"→"Smith") were added
as reveal pairs, causing partial replacements that corrupted the DOM.
The smart engine sends combined forms ("Ademo Demo"→"John Smith")
which the catch-all already handles. Removed individual name entries
from buildRevealPairs — only emails, usernames, hostnames, phones
are matched individually.
2. Cache invalidated on every ss:config-updated (including settings-only
changes like reveal toggle). Now only invalidates when mappings or
identity actually change.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
If reveal mode was saved as ON in settings, the page loaded with
revealMode=true and prevRevealMode=true. The checkRevealToggle
function only fires on transitions (off→on or on→off), so the
reveal interval was never started. Added initialization check
that starts the reveal interval immediately if revealMode is
already true at page load.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
AMO expects { required: false } not { collect_user_data: false }.
Bumped strict_min_version to 140.0 (when data_collection_permissions
was introduced). Version 0.9.8.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
When at-rest encryption is enabled, storage.onChanged fires with
encrypted blobs for ss_settings. The injector was passing this
encrypted blob directly as settings to the page world content script,
overwriting real settings with { _ssLocalEncrypted: true, data: ... }.
This broke reveal mode, highlights, and any setting toggle because
the content script's settings object became the encrypted blob.
Fix: skip encrypted settings blobs in injector.js (same check already
existed for mappings and identity). The background's settings:updated
message already sends decrypted settings correctly.
https://claude.ai/code/session_01KF4i7Ra7zCEDskxDBaNtcT
Claude