Every day I use Claude.ai for real work — writing specs, refining prompts, making product decisions.
And every single day, the same painful ritual before I could do any of it:
New chat → explain who I am → explain my project → explain my tone → explain my context → okay, now we can actually start.
I got tired of it. So I built something.
The 3 things silently killing my Claude workflow Claude forgets everything, every session.
No memory across chats. You're starting from zero every time. If you use Claude seriously, you already know this pain.
No visibility into token usage.
Claude Pro has a context window limit — but no meter, no warning, nothing. Your conversation quietly degrades mid-workflow and you're left wondering why the outputs suddenly got worse.
Most prompts are terrible. Mine were too.
Vague prompt → vague output. Rewriting prompts properly takes time and skill most people don't have in the middle of actual work.
What I built: Promptive Sentry
A free Chrome extension that fixes all three — right inside Claude.ai.
🧠 Sentry Memory — Save your context once. Load it into any new conversation in one click. Your role, your project, your instructions — always ready.
📊 Token Monitor — Real-time usage tracker inside Claude.ai. Know exactly where you are before quality starts dropping.
⚡ Prompt Upgrader — Paste your rough prompt, get a structured version back. Better input, better output.
No account needed. No data stored. Completely free.
The thing that surprised me most
When I showed early users the memory feature, almost everyone said the same thing:
"Why doesn't Claude just have this built in?"
That gap between what AI tools could do natively and what they actually do for real workflows — that's where I'm building.
👉 Install it here — free, no strings attached.
https://chromewebstore.google.com/detail/promptive-sentry-%E2%80%93-author/ikbkijdgnelcijmdkcaaoabmobagakim
I'm building in public and shipping more tools soon. What's your biggest frustration with Claude or ChatGPT right now? Genuinely curious.
