FAQ from PostPerfect
What is PostPerfect?
PostPerfect is a purpose-built AI enhancement layer for Twitter—designed to refine clarity, boost engagement, and express intent more precisely—all while preserving your authentic voice.
How to use PostPerfect?
Install the extension, connect via Twitter’s official OAuth flow, pick a style (e.g., “Concise Analyst,” “Witty Engager,” or “Empathetic Advocate”), and start typing. Suggestions appear live—edit or accept in-context. Done.
How does PostPerfect work?
It injects a lightweight, isolated script into twitter.com that observes only your active compose field—not your timeline, DMs, or history. Using local inference powered by an optimized GPT-3.5 Turbo pipeline, it generates stylistically coherent alternatives without sending prompts to external servers.
Can I add my own styles to PostPerfect?
Absolutely. Define custom style cards using plain-language instructions (“Be concise, use active verbs, avoid jargon”)—then save, rename, and reuse them across sessions. Your personal style library stays encrypted and local.
Can I use PostPerfect on mobile devices?
Not yet. It’s currently optimized for desktop Chrome and Chromium-based browsers (Edge, Brave, Arc, Vivaldi) to ensure full compatibility with Twitter’s dynamic interface and robust security controls.
Does PostPerfect track anything?
No tracking—ever. No analytics, no telemetry, no keystroke logging. Your tweets, styles, and usage patterns never leave your browser. We even disable network requests during suggestion generation.
Why are my suggestions not perfect sometimes?
AI augmentation is iterative—not absolute. While PostPerfect leverages state-of-the-art language understanding, nuance depends on input clarity and stylistic specificity. We continuously refine our system prompts and fine-tune response logic based on anonymized opt-in feedback.
What social media platforms are supported?
Today, PostPerfect is engineered end-to-end for Twitter (X.com). LinkedIn and Reddit integrations are in active development—and will follow the same privacy-first, in-browser architecture. Facebook support is under evaluation.