Communication Analysis & Context Interpretation Tool
Decipher-It is a single-file browser application designed to analyse phrases, described tone and user-provided context. It combines offline pattern matching with optional online AI analysis, offering confidence indicators, risk tags, analysis history, diagnostics and crash recovery.
- Three-pillar analysis: Linguistic, paralinguistic and contextual layers.
- Dual-mode operation: Offline phrase-pattern engine and optional online AI providers.
- Four tabs: Analyse, History, Settings and About.
- Risk indicators: Confidence bar, risk-factor tags and colour-coded danger levels from 1–5.
- Analysis history: Stores the last 50 analyses in browser storage when available.
- Voice input: Browser-based speech recognition where supported.
- Crash recovery: Automatic retries, diagnostics, recovery snapshots and safe mode.
- Responsive interface: Designed for desktop and mobile browser layouts.
- Linguistic Layer: Examines the words and phrases provided.
- Paralinguistic Layer: Uses tone, pitch and tempo descriptions supplied by the user.
- Contextual Layer: Considers the background and history provided by the user.
The tool offers possible interpretations, not definitive conclusions about another person's intentions or psychological state.
- Offline Mode: Uses rule-based pattern matching for selected phrases and expressions. It does not require an API key.
- Online Mode: Connects to optional AI providers when configured with a valid API key and supported model.
- Automatic Fallback: Returns to offline analysis when online requests fail.
- Diagnostics: Checks connectivity and helps identify configuration or request errors.
- Privacy: Online analysis sends submitted content to the selected provider. API keys are entered in Setup and stored in the browser; review provider policies before submitting sensitive information.
- Use the file
decipher-it-v5.html. - Open it in a modern web browser.
- Use offline analysis without configuring an API key.
- For online AI, open Settings/Setup and configure a supported provider.
- For microphone access, host the file over HTTPS and grant browser permission.
Recommended: Keep a backup of the HTML file and use the v5 file rather than superseded v4 or JSX drafts.
- 73 of 73 browser checks passed in headless Chromium.
- Tested with desktop and phone-sized browser emulation.
- Checks cover analysis, disclaimer gate, crash recovery, diagnostics, corrupted-data repair, safe mode, microphone flow and simulated AI responses.
- Live AI calls, current provider model IDs and real API-key connectivity remain unverified.
- Speech recognition uses browser speech services and requires internet; it is not fully offline.
- Real iPhone, Android and Windows hardware testing remains outstanding.
- SOS and crisis-support entries marked for verification must be checked against official sources.
- The offline engine is rule-based, not an on-device language model.
- The tool is for communication interpretation and should not replace professional advice, direct communication or emergency services.