Introducing Reverse Prompt, a powerful new feature in PrompTessor that lets you uncover the prompts behind any content — from images, text, or URLs. Simply upload an image, paste text, or share a link, and PrompTessor instantly analyzes it to generate possible prompts that could have created the same output. What’s new in this release: 🧠 Reverse Prompt — analyze any image, text, or URL to discover the prompts behind it. ⚙️ Refine with Feedback integration — refine reversed prompts instantly with contextual feedback (e.g., make it Midjourney-ready). ⚡ Improved prompt analysis engine — faster and more accurate prompt reconstruction powered by enhanced content recognition. 🪄 Optimized UI/UX flow — smoother transitions between Reverse Prompt, Refine, and Optimize tools. 🔍 Better prompt output clarity — refined formatting for cleaner, more readable prompt results. 🚀 Performance & stability improvements — faster response time and reduced latency during prompt generation. 💎 Lifetime Deal (limited spots) — get lifetime full access to all current and future features of PrompTessor at a one-time price, available only for a short time. Reverse Prompt transforms how creators, prompt engineers, and AI enthusiasts understand, learn, and recreate creative outputs

Hi there 👋 Flo here — this one’s a banger: a brand‑new agentic engine, 30+ new integrations — including Google Calendar, Gmail and Drive and long‑term memory. Everything bellow is released today in BETA. If you want to join the new beta program [you’ll first need to join the Slack community](https://join.slack.com/t/mindtheflo/shared_invite/zt-3bk60o4on-MnEfNtUfT_FYq2_w7EXSew) and you’ll find the instructions under the [beta Slack channel](https://mindtheflo.slack.com/archives/C09KRDWCG3A). Starting next month, the Beta channel will be reserved for Pro+ users as will be most of the features bellow; **so now is your chance to activate those features for free!** Ok, let’s roll. New agentic engine - Notis V3 Notis’ engine was rebuilt on OpenAI’s latest Responses/Conversations API stack for significant gains in speed, reliability, and intelligence. Notis now replies instantly, then smart‑routes work to specialized agents with the right “brainpower” for the job. Concretely Flo? - Expect the first response in seconds not minutes. - Simple tasks/automations are cheaper and faster. - Much improved intelligence and autonomy on complex tasks/automations. Email, Calendar, Drive and 33+ integrations (PRO+) Without a doubt, the number one requested feature since I launched Notis 12 months ago has been giving access to Notis to your calendar. Well, today I am releasing 33 integrations in Beta. Concretely, Flo? - You can draft/send emails from Notis or review your inbox during your morning job. - Use Todoist or any other task manager to manage your tasks from Notis. - Work on your sales pipeline in HubSpot or Salesforce. - Report bugs on GitHub, Linear, or Jira. - Add a blocker in your agenda for the time you need to go to your meeting (Google Maps + Calendar). You get the gist of it… You can edit your integrations in the Integrations page in the portal or simply asking Notis I believe we now have the architecture and resources to move extremely fast to deliver the integrations that matter to your productivity. [Now, go vote for the next integration here!](https://notis.featurebase.app/) Long-term memory (PRO+) Right next to Google Calendar, long-term memory was the underlying request behind much of the users’ feedback. And here it is. Why wait so long? I was not convinced of the solution until now. Notis long-term memory is managed automatically: - It’s used for context for every interaction with Notis. - It’s updated continuously with every new important piece of information Notis learns about you. No user guide. No tips. It works like magic. 🐞 Other small improvements Most of these would deserve their own header image, to be honest, but I limit myself to three new major features a month! - Priority and Flex service tier: Automations now run for half the cost by using the OpenAI Flex tier. You can opt in for the Priority tier in your user settings for 2x the base token costs but even faster Notis. - Code interpreter: Notis now spins up a sandbox to run code, perform complex calculations, or analyze files. It supports working on documents, so for example, you can easily use Notis to analyze a CSV file. - Improved logging: Detailed agent traces in Portal → Usage. Also, easily filter your automation logs in Portal → Automation → Logs. Logs now also show the automation that triggered it, the source, and the channel it ran on. - Automation improvement: Name automations for clarity and the option to run them in a fresh context per run. - Notion improvements: Notis can now edit Notion covers/icons and better handle images. - Deep Research improvements: Notis can now conduct deep research on your Notion workspace and any attached docs. - Scraping tool improvements: The tool now captures page screenshots during scraping. That's it for this month. See you next month! Flo PS: This is the first changelog entirely drafted by Notis and sent right from Notion. Notis is now doing the job of 2 to 3 full-time employees — this is how good it has become.

Hey Athena users! Sora 2 is now on Athena! Generate high-quality videos! Unleash your creativity with the power of Sora 2 and Veo 3! You can now generate stunning, high-quality videos directly within Athena, up to 12 seconds in length. Here’s how to use it: Simply tell Athena the content of your video. Important: You must specify both the desired format (Landscape or Portrait) and the length (4, 8, or 12 seconds) in your prompt. Example: "Make video of princess having a dance in a ballroom with her dog! has to be 12 seconds long and landscape format."

What's new • 10x speed up in company search, without loss of quality • People search AI agent that verifies person currently works at a company • Waterfall people contact enrichment across 20+ providers (powered by FullEnrich)

We've completely redesigned Alice for the best AI experience on desktop. Version 4 gives you superpowers unlike any other AI app. We're also introducing single, tiny subscription starting from $8/mo including both App and AI tokens for the latest models! Key features of Alice 4.0 are: ✓ integration with all major providers (BYOK) ✓ subscription access to state-of-the-art models ✓ fully private workflows with ollama and lmstudio ✓ profiles (assistants) and prompts (skills) ✓ global keyboard shortcuts for skills ✓ configurable in-app shortcuts ✓ assistants Library and Assistants Generator ✓ thread splitting ✓ compact and extended UI modes ✓ system notifications and sound effects ✓ advanced and finely configured infinite conversation history ✓ native web search for all major providers or firecrawl ✓ audio integration (TTS / STT) ✓ image generation with Nano Banana and GPT-Image-1 ✓ document organization in folders and assistants, using file exploration instead of basic vector search ✓ memory features for basic facts and instructions ✓ MCP support for tools, prompts, resources, and sampling through STDIO, SSE, and streamable HTTP, including OAuth ✓ MCP tool organization within assistants ✓ augmented MCP where skills and memories influence how MCP is used ✓ semi-advanced model settings ✓ custom connections to your own server that supports the OpenAI API And many more 🎉

Introducing Auto Organization. We’ve all been there: a cluttered drive full of random files we swear we’ll organize “later.” Now you don’t have to. Every file you upload is automatically sorted into the right folder — or new ones are created for you. Just drag, drop, and watch your workspace organize itself.

- Added visible record/pause indicator. - Improved filler-word cleanup (“ums/uhs”) for more polished transcripts. - Publish a clear privacy & security page (GDPR, data retention, where data lives).

Between you and your vision lies a mountain of problems. Apple, SpaceX, Tesla. They all faced impossible odds. A mountain of problems nobody knew how to solve. What separated them wasn't genius. It was the relentless ability to solve one problem after another. What is unstoppable momentum? It's when solving problems feels effortless. You're not stuck or confused. Each step forward feels natural, like you're in flow. Progress compounds because nothing gets lost. Every insight builds on the last. Every solution creates knowledge that makes the next problem easier. How socra works: 1. Clearly define the problem Frame the actual problem, not the symptoms 2. Break big problems down Break down complex problems into solvable steps 3. Find the actual root cause Dig past symptoms to understand what's actually broken 4. Work on solutions together Collaborate with people and AI to find better solutions faster 5. Build your knowledge Every solution creates knowledge that makes the next problem easier Why socra? Memory without limits socra builds memory over time. Creating deep understanding. Evolving with every interaction. Making starting over obsolete. Collaboration reimagined Solve problems together with persistent memory, clear ownership, and structured thinking. Not just chat. Not just comments. Real collaboration that builds knowledge. AI you can trust Built for deep, meaningful collaboration. Your thoughts stay yours. Private, secure, and within your control. Different ways of thinking Work with AI agents modeled after Einstein, Jobs, Turing, and more. Each applies a different thinking and perspectives to help you break through. Start your socra journey today See why 70,000+ people trust socra to achieve more, together. Compound your knowledge. Amplify your thinking. Multiply your achievements.