Outgoing Message Blocking
Every other content filter focuses on what comes in. Peace of Mind is the only filter that stops what goes out. Type a blocked word in a DM, a chat box, a search bar, or any text field on any website and hit Send. The message never leaves your device. The server never receives it. The recipient never sees it.
This matters because seeking behavior almost always starts with typing. A search query. A message to someone. A username in a search box. By the time the results come back, the damage is already done. Peace of Mind stops the chain at the very first link — the moment your fingers hit the keys.
No other content filter does this. Covenant Eyes, Net Nanny, Qustodio, BlockerX — they all let the message leave your device and then try to deal with what comes back. Peace of Mind prevents the request from ever being made.
Context-Aware Intelligence
Most filters treat every website the same. A word is either blocked everywhere or blocked nowhere. This forces you to choose between crippling your productivity or leaving gaps in your protection. Peace of Mind understands context.
Writing about your recovery in Notion? Not blocked. Processing your triggers in Google Docs? Not blocked. Journaling in Standard Notes or Evernote? Not blocked. The same words that are perfectly fine in a recovery journal are caught and blocked instantly on Instagram, Discord, Reddit, and every other platform where they signal seeking behavior.
This is the difference between a filter that fights you and a filter that works with you. Your recovery writing is sacred. Peace of Mind treats it that way.
Real-Time Input Scanning
Peace of Mind does not wait for you to press Enter. It scans every text field on every page in real-time, keystroke by keystroke. The moment a blocked word is formed in any input field — a search bar, a chat box, a comment field, a contenteditable editor — the page is blocked before the search even starts.
This is fundamentally different from URL-based or DNS-based filtering. Those tools wait until the harmful content arrives. Peace of Mind intervenes at the point of intent, before the request is made, before any content is fetched, before anything reaches your screen.
Every input field. Every page. Every keystroke. In real-time.
13M+ Domain Bloom Filter
Peace of Mind maintains the largest domain blocklist of any content filter — over 13 million domains. The entire list is checked in microseconds using a bloom filter, a data structure that delivers instant lookup with zero false negatives. If a domain is on the list, it is blocked. Every time. Without exception.
The entire bloom filter runs on your device. No server lookup. No DNS delay. No third-party service that could go down or be bypassed. The domain is checked before the page even begins to load, and the result is available in less time than it takes your screen to refresh.
New domains are added regularly. The blocklist is compiled from multiple independent sources, merged, deduplicated, and compressed into a single binary that takes up just 37MB on your device.
Full Unicode Normalization
Most filters check against a simple list of English words. Creative spelling defeats them in seconds. Peace of Mind normalizes every piece of text through a comprehensive pipeline before checking it: Cyrillic lookalikes converted to Latin (the Cyrillic "a" is not the Latin "a" — Peace of Mind catches both), leetspeak decoded (3 becomes e, @ becomes a, $ becomes s), emoji substitutions resolved, zero-width characters stripped, diacritical marks removed, soft hyphens collapsed.
This means every creative evasion technique that works against other filters — homoglyphs, invisible characters, mixed-script text, Unicode tricks — fails silently against Peace of Mind. The text is normalized before it is ever checked, so every variant maps back to the same underlying word.
Users cannot trick the filter with creative character substitution. The pipeline handles every known Unicode evasion technique in the wild.
Fuzzy Matching
Even after Unicode normalization, intentional misspellings remain a common bypass strategy. Peace of Mind handles this with fuzzy matching: missing letters, doubled characters, adjacent character swaps, and common phonetic substitutions are all caught automatically.
The system checks over 220,000 word variants in under 0.05 milliseconds. This is fast enough to run on every text node on every page without any perceptible delay. You will never notice it working — but every intentional misspelling, every creative respelling, every attempt to sneak a word past the filter by changing a letter or two is caught and blocked.
Combined with Unicode normalization, this makes Peace of Mind's text scanning the most thorough of any content filter available. No other product matches the depth or speed of this analysis.
Pre-Blur Image Scanning
Peace of Mind does not scan an image and then decide whether to blur it. It blurs every image first, before anything renders on the page. A 40px Gaussian blur is applied via CSS injected at document start — before the browser has drawn a single pixel. You never see harmful content, not even for a split second.
Once the page loads, an on-device neural network (InceptionV3) classifies each image into five categories. Safe images are cleared and the blur is removed instantly. Suggestive images stay blurred. Explicit images trigger a full page block. The entire process runs locally on your device — no images are ever sent to any server.
This blur-first approach is the opposite of what other tools do. Covenant Eyes and similar products see the image, analyze it, and then report it after the fact. By that point, the image is already in your head. Peace of Mind ensures you never see it in the first place.
Network Traffic Interception
Modern web apps load content dynamically. Instagram loads posts through background API calls. Twitter fetches tweets via XHR requests. Discord sends messages over WebSocket connections. None of this content appears in the URL bar, and none of it is visible to DNS-level filters.
Peace of Mind intercepts every fetch request, every XHR call, and every WebSocket message. It scans both incoming responses and outgoing request bodies. This means content that never appears in the visible page — content loaded asynchronously, content fetched in the background, content streamed through persistent connections — is still caught and blocked.
This is how Peace of Mind protects you on social media platforms where every other content filter fails. The content is intercepted at the network level before it ever reaches the rendering engine.
Tamper-Resistant Lockdowns
When you activate a panic lockdown, trigger a cool-off period, or enable a social media blackout, the protection persists even if someone clears all extension data, wipes browser storage, or attempts to reset the extension. The blocking rules operate at the browser's declarativeNetRequest level, independent of extension storage.
This means a moment of weakness cannot undo your protections. Clearing cookies does not end a lockdown. Deleting local storage does not cancel a cool-off. Restarting the browser does not disable a blackout. The rules are registered at the network layer and persist until their timer expires — no exceptions, no workarounds, no escape hatch.
Combined with the lockdown guide (which disables developer tools, blocks incognito mode, and prevents extension removal), this creates a system where the filter cannot be circumvented through any standard browser mechanism.
Impulse Resistance by Design
Every content filter lets you weaken your own protections in a moment of weakness. Peace of Mind does not. The core design principle is simple: adding protection is instant, removing protection takes three days.
Block a site? Instant. Unblock it? Three-day waiting period. Enable a social media restriction? Instant. Disable it? Three-day waiting period. Add a word to the blocklist? Instant. Remove it? Three days. Every single change that would weaken your protection has a mandatory cooling-off period.
This is not a bug or an inconvenience — it is the entire point. Addiction exploits impulsive decisions made in moments of weakness. Peace of Mind makes it structurally impossible for a moment of weakness to change anything. By the time the three days are up, the urge has passed.
Video Protection — 4 Layers
Video content requires protection beyond static image scanning. Peace of Mind uses four independent layers to handle video across every platform. Thumbnail propagation: if a video thumbnail is flagged, the video itself is blurred without needing to analyze video frames. Hover preview detection: YouTube and other platforms show preview clips on hover — Peace of Mind catches and blurs these before they play.
Soft block video freeze: when a soft block is triggered on a page, all video playback is paused immediately — no explicit video continues playing while you read a warning. DOM propagation: when new video elements appear near already-flagged content (infinite scroll, dynamically loaded feeds), they are caught and blurred automatically by the mutation observer.
These four layers work together to ensure video content is handled as thoroughly as static images, across YouTube, Twitter, Reddit, Instagram, and every other platform that serves video.
Smart Soft Block
A single concerning word in a page of otherwise safe content should not lock you out entirely. That one spam email in Gmail, that one old message thread in Discord, that one comment on an otherwise safe Reddit post — a hard block on every page with a single match would make the internet unusable.
Peace of Mind uses a tiered response. One or two unique concerning words trigger a soft block: all media on the page is blurred, all videos are paused, and a banner appears — but the page stays usable. You can still read your email, still use your chat app, still browse the thread. Three or more unique words, or a match in the page title, triggers a full page redirect.
This graduated response keeps you protected without punishing you for encountering a single word in otherwise safe content. The filter is strict where it needs to be and proportional everywhere else.
Bypass Tool Blocking — ~375 Domains
The most common bypass for any content filter is downloading a different tool: a VPN to tunnel around DNS restrictions, an alternative browser without the filter installed, a proxy to route traffic around blocks, a remote desktop app to access an unfiltered machine. Peace of Mind blocks the download pages for all of these — approximately 375 domains and paths.
The blocking is path-specific where it needs to be. Brave Search is accessible because it is a legitimate search engine, but the Brave Browser download page is blocked. Mozilla's MDN documentation is accessible, but Firefox download pages are blocked. Discord's web app works, but the Discord desktop app download is blocked. GitHub is accessible, but specific repositories known to distribute bypass tools are blocked.
This surgical approach keeps legitimate tools accessible while closing every known route to downloading software that could circumvent the filter.
AI Image Classification — InceptionV3
Peace of Mind uses InceptionV3, a neural network originally developed by Google, to classify every image on every page into five categories. The model runs entirely on your device — no images are ever uploaded to any external server. Your browsing remains completely private.
The system uses a two-tier response calibrated through extensive real-world testing. Suggestive images are blurred so you can still navigate the page. Explicitly harmful images trigger a full page block. The thresholds are tuned to minimize false positives on normal content while catching harmful content with extremely high accuracy.
This is the same class of neural network used in commercial content moderation systems, running locally on your device at the speed of your browser. Combined with the pre-blur system that blurs every image before the AI has finished classifying it, this creates a pipeline where harmful images are never visible — not while loading, not while being classified, not ever.