# Invoices & payment Source: https://help.notis.ai/account-and-settings/billing/invoices-and-payment Manage your subscription, find invoices, update payment details, cancel, and request refunds. ### How do I access billing settings? Open the [Notis portal](https://app.notis.ai/usage) and click **Manage Subscription** in the right-hand sidebar of the Usage page (or in the user menu). This opens a secure Stripe session for your account where you can update payment methods, switch plans, download invoices, and cancel. If you prefer the email-based flow, open the [Stripe customer portal](https://billing.stripe.com/p/login/3cs3e95a76jl1pe144) directly and enter the email you used when subscribing — you'll receive a secure link by email. ### Where can I find my invoices? Open the [Notis portal](https://app.notis.ai/usage), click **Manage Subscription**, and scroll to the **Invoice history** section in the Stripe billing portal. You can open the hosted invoice page or download a PDF for every current and past invoice, including On-Demand usage settlements. Each On-Demand invoice is for the exact overage balance accumulated when it is settled. Automatic On-Demand payments are grouped to at most one successful charge every 24 hours, so new usage during that window appears on a later invoice instead of creating several small charges. Invoices are also emailed to your billing email each time a payment succeeds. ### How do I update my billing information? Open the [Notis portal](https://app.notis.ai/usage) → **Manage Subscription** to update your payment method, company name, address, and tax information through Stripe. All transactions are processed securely by Stripe. Changes only affect future invoices — historical invoices cannot be modified. ### How do I cancel my subscription? You can cancel in two ways: 1. **Ask Notis directly** in any channel ("cancel my subscription"). 2. Or cancel yourself: open the [Notis portal](https://app.notis.ai/usage), click **Manage Subscription**, then click **Cancel subscription**. Your access to Notis continues until the end of your current billing period, and you can keep using your remaining usage cap until that date. ### How do I request a refund? You can request a refund **directly from Notis** by asking in any supported channels. If needed, our support team may ask for a few details (like the email on the subscription and the invoice) to locate the payment quickly. ### I'm having other billing issues. How can I get help? For billing questions not covered here, use our live chat in the bottom right of the help portal or ask Notis for Flo's contact details. # On-Demand Usage Source: https://help.notis.ai/account-and-settings/billing/on-demand-usage Keep working past your plan's monthly cap and pay only for the overage. On-Demand is currently in **beta**. The behavior described below may change as we refine pricing and limits. On-Demand is an opt-in overage layer for paid subscriptions. Your plan's included usage is consumed first, followed by any usage credit in your wallet. Once both are exhausted, On-Demand lets Notis keep running and bills the extra usage separately. Think of it like a safety valve: instead of stopping mid-conversation when your included credits run out, Notis can continue as long as On-Demand is enabled, your payment method is valid, and you have not reached the On-Demand limit you set. ### Who can use it? On-Demand is available to any paid subscriber with a payment method on file. Free trials and users without an active subscription cannot enable it. If you're on a team plan, only the **team owner** can change On-Demand settings. Team members inherit the owner's configuration as read-only. Everyone on the team uses the same included usage pool and the same On-Demand limit. ### How does it work? 1. Notis consumes your included monthly usage and any wallet credit first — On-Demand stays idle. 2. When both are exhausted, On-Demand kicks in and Notis keeps responding. 3. Overage is tracked against your monthly On-Demand limit. On teams, this limit applies to the whole team, not each teammate. 4. Notis periodically invoices the accumulated overage balance once it reaches your settlement threshold. Automatic payments are grouped so there is at most one successful On-Demand charge every 24 hours for you or your team. 5. At the end of the billing cycle, Notis charges any remaining overage balance if it is large enough for Stripe to process. Tiny balances carry forward and are included in the next settlement. You can monitor your overage usage and unbilled balance from the **On-Demand** section on the [subscribe page](https://app.notis.ai/subscribe). ### The three modes When you open the On-Demand card, you choose one of three modes: * **Fixed** — Set a hard monthly USD ceiling (for example, \$20). The task that reaches the remaining limit can finish, but you are never billed above the limit. Later tasks pause until you raise the limit or a new billing cycle begins. * **Unlimited** — No monthly overage ceiling. Notis keeps working as long as your payment method stays valid. Use this if you'd rather not be interrupted and trust your team to self-regulate. * **Disabled** — Default. Once your included usage is exhausted, Notis pauses until your billing cycle resets or you upgrade. Start with a **Fixed** limit that matches roughly one extra month of plan usage. You can switch to Unlimited later once you have a feel for your overage pattern. ### How am I billed? On-Demand is billed in arrears — you only pay for the extra model usage you actually use. * Each overage request adds to your unbilled balance using the same Notis model rates listed in [Usage & model rates](/account-and-settings/billing/usage-and-rates). * When your unbilled balance reaches your settlement threshold, Notis automatically charges the exact outstanding overage amount. * To reduce small card charges, successful automatic settlements are grouped to no more than one every 24 hours. Your exact new usage keeps accumulating for the next invoice during that window. * At the end of each billing cycle, Notis tries to settle any remaining overage balance before your usage counter resets. * If the remaining balance is too small for Stripe to charge, it carries into the next cycle and is included in your next settlement. * Carried overage does **not** reduce your next month's included plan credits. Your first settlement threshold is \$1.00. As you build a good payment history, Notis can wait longer before charging so you see fewer small card charges: | Good-standing history | Settlement threshold | | --------------------------------------- | -------------------: | | New On-Demand account | \$1 | | 1 successful On-Demand charge | \$3 | | 3 successful On-Demand charges | \$5 | | 10 successful On-Demand charges | \$10 | | 10 successful charges and 60+ paid days | \$20 | If an On-Demand payment fails, the threshold resets to \$1.00 until the payment issue is fixed. If you use a **Fixed** monthly limit, that limit controls whether new overage can continue; it does not change the settlement threshold. For example, if your threshold is \$5 and your overage balance is \$3.20, Notis keeps tracking the balance but does not charge yet. If your next task brings the balance to \$5.40, Notis charges \$5.40. If your billing cycle ends with only \$0.30 outstanding, that \$0.30 carries forward and your included plan credits still reset in full. On-Demand charges are processed securely through Stripe as invoices. Each invoice shows the exact accumulated amount being settled and includes a hosted invoice page and downloadable PDF. You can find On-Demand and subscription invoices from the [Notis portal](https://app.notis.ai/usage) → **Manage Subscription** → **Invoice history**. ### What happens if a payment fails? If Stripe can't charge your card for an On-Demand settlement, your On-Demand status switches to **suspended**: * Further overage usage is blocked. * Your plan's included usage still works normally until it's exhausted. * A red banner appears on the On-Demand card. If you're the subscription owner, the banner can show: * **Retry payment** when there is an overage balance ready to charge. * **Resume on-demand** when the remaining balance is too small to charge right now. * **Update payment** when you need to change the card in Stripe. After you update your payment method and return to Notis, Notis automatically retries the same open invoice for you. You can also click **Retry payment** from the On-Demand card. A failed retry does not create another invoice for the same balance. If the balance is too small for Stripe to charge, Notis can resume On-Demand and carry that small balance forward. It will be included the next time your overage balance reaches the settlement amount. If you're a team member, ask your team owner to update payment or retry recovery. Team members can view the team's On-Demand settings, but only the owner can change or recover them. ### Enabling On-Demand 1. Open the [subscribe page](https://app.notis.ai/subscribe) in your portal. 2. Scroll to the **On-Demand** section below the pricing grid. 3. Pick **Fixed** (and set a monthly limit) or **Unlimited**. 4. Click **Save**. To turn it off, switch the mode back to **Disabled** and save. You'll keep access to any unbilled balance through the end of the current cycle, and no new overage will be charged after the change takes effect. ### FAQ **Does On-Demand reset every month?** Yes. Your On-Demand usage counter resets at the same time as your plan's billing cycle. **Does carried debt reduce next month's plan credits?** No. If a sub-minimum balance carries forward, it's added to your next settlement — your included plan credits reset in full. **Can I cap a teammate's On-Demand spend?** On-Demand limits apply to the whole team subscription, not per-member. Use a **Fixed** limit to bound total team overage. **Do team credits pool together?** Yes. A team's included usage is shared. If your team has 10 paid seats and each seat includes \$150 of monthly usage, the team gets one \$1,500 monthly pool. **Does the 7-day Ultra trial include On-Demand?** No. On-Demand is only available after you activate a paid subscription. # Plans & trial Source: https://help.notis.ai/account-and-settings/billing/plans-and-trial The Notis tiers, the 7-day free trial, and how upgrades, downgrades, and billing-cycle switches work. Notis offers flexible pricing tailored to how much value you're getting from your AI intern, with both monthly and annual billing options. Annual plans get **4 months free**. ### What are the different tiers? * **Pro** (\$20/mo or \$13/mo when billed annually): For solo founders and operators who want the convenience of a personal PA in WhatsApp, iMessage, Telegram, and email. * **Pro+** (\$59/mo or \$39/mo when billed annually): For busy operators who need to win back a full day every week with skills, automations, and Notis Apps. * **Ultra** (\$149/mo or \$99/mo when billed annually): For high-growth teams building human-led, agentic teams—where operators design and manage a swarm of AI agents. Ultra includes [Your Computer](/your-computer/overview), so Notis can work with your local apps and files after you activate it in the desktop app. ### Do you offer a free trial? All plans come with a 7-day free trial with a \$20 usage cap. You enter your payment details upfront, and your chosen subscription starts automatically at the end of the trial unless you cancel. The trial is for new accounts: it's available during your first week with Notis, and once per account. If your account is older than that or you've already used a trial, the subscribe page won't show the trial offer and your plan starts right away when you subscribe. The trial gives you exactly the plan you picked: trialing Pro+ includes skills, automations, and Notis Apps; trialing Ultra adds [Your Computer](/your-computer/overview), memory connectors, [Premium Tools](/tools/included-tools), and the other Ultra features. If you want to try a feature from a higher plan, switch your trial to that plan at any time — just ask Notis or pick the plan on the [subscribe page](https://app.notis.ai/subscribe). Switching during the trial is free and keeps your original trial end date. Store browsing remains available on every signed-in plan. You can activate your plan at any time during the trial to unlock the full usage cap — just ask Notis or click **Activate now** on the [Usage page](https://app.notis.ai/usage) of the portal. ### Will I know before I run out of usage? Yes. When you get close to your monthly usage cap, Notis adds a heads-up to its next reply so you're never caught off guard. That gives you time to raise your limit on the [Usage page](https://app.notis.ai/usage) before anything pauses. ### What are Notis's billing cycles? Billing cycles run monthly or annually, starting on the day you subscribed. Subscriptions renew automatically unless canceled before the renewal date. You can see your current cycle on the [Usage page](https://app.notis.ai/usage) under **Billing cycle**. ### Can I switch between monthly and annual billing? Yes, you can switch self-serve from the Notis portal: 1. Open [app.notis.ai/subscribe](https://app.notis.ai/subscribe). 2. Toggle **Annual billing** on or off at the top of the pricing page (annual gets you 4 months free). 3. Pick your plan and confirm in Stripe Checkout. The change takes effect immediately. Switching from monthly to annual on the same tier charges the prorated difference; switching from annual to monthly applies at the end of your current billing period so you keep the months you've already paid for. ### How does Notis handle billing changes? When you change your subscription plan or billing cycle, Notis follows two simple rules: * **Upgrades take effect immediately.** Moving to a higher tier (or from monthly to annual on the same tier) charges the prorated difference right away and unlocks the new usage cap on the spot. * **Downgrades take effect at the end of the period.** Moving to a lower tier (or from annual to monthly) keeps your current plan and usage cap until your billing period ends, then switches you over — so you don't lose anything you've already paid for. ### What happens to my reminders and automations if my subscription ends? If your subscription lapses, Notis automatically pauses your reminders and automations so you stop receiving messages you're no longer subscribed for — you don't have to turn them off one by one. When you resubscribe, the reminders and automations that were paused for this reason turn back on automatically (automations return once your plan includes automations). Anything you paused or deleted yourself stays that way. # Usage & model rates Source: https://help.notis.ai/account-and-settings/billing/usage-and-rates How your monthly usage budget is tracked, what each model costs, and what happens when you reach the cap. ### How do limits work? Each plan includes a monthly agent usage budget denominated in dollars. Each Notis message deducts from that budget using Notis model rates, which include a 20% service markup over the provider's published API price for OpenAI models. You can view your billing cycle, included usage, On-Demand usage, and request breakdowns on the [Usage page](https://app.notis.ai/usage) of your portal. In a Manager thread, the usage indicator also shows the thread's completed cost and how much context the latest completed run used. If you use Notis with a team, everyone shares one monthly usage pool. For example, a 10-person team on a plan with \$150 of included monthly usage per seat gets a shared \$1,500 pool for the month. Any teammate can use it, and all team activity counts toward the same total. The Usage page shows the team's shared usage for the current billing cycle, and billing settings are managed by the team owner. Notis manages model routing automatically based on the task. Prices below are shown per 1M tokens and reflect the current Notis rates after markup. ### Auto mode By default, Notis runs on **Auto** (recommended). Instead of a fixed model, Notis's smart router reads each turn and picks the intelligence level — model and reasoning effort together — that fits the task: light, inexpensive work for a quick reply, and a stronger model for a hard, multi-step task. Because the choice is made per turn, a single conversation can move up and down levels as the work changes, so you only pay for the capability each step actually needs. ### Intelligence levels
PRO+ Pin custom defaults with Pro+ and Ultra
You can **pin** an intelligence level — per conversation in the [Desktop and Web app](/channels/manager#intelligence-and-priority), as a per-channel default in [Channels](/channels/channels-compared#default-intelligence-%26-priority), or per automation in the [automation editor](/get-started/automate#intelligence) — instead of letting Auto choose. Each level sets the model and its reasoning effort together: | Level | Model | Reasoning effort | Price vs Medium | | ------ | ------------- | ---------------- | --------------: | | High | gpt-5.6-sol | Medium | ×2.5 | | Medium | gpt-5.6-terra | Low | ×1 | | Low | gpt-5.6-luna | XHigh | ÷10 | The multiplier compares each model's regular per-token rates with Medium. Reasoning can change how many tokens a message uses, so actual billing always uses the real token counts and rates below. ### Processing tiers (priority) Notis chooses the processing tier automatically, but on a plan with priority access you can pin one per thread or per channel alongside the intelligence level (it shows as **Standard / Fast / Economy** in the picker): | Tier (picker label) | Price vs Standard | When it's used automatically | | ------------------- | ----------------: | ------------------------------------------------------------------ | | Fast | ×2 for GPT-5.6 | Voice interactions when the model supports it | | Standard (Normal) | ×1 | Default for regular messages and user-triggered work | | Economy (Flex) | ÷2 | Background or automation-triggered work when the model supports it | If a model does not support Economy or Fast, Notis uses its Standard rate instead. The Economy and Fast token tables below show the resulting per-token rates. **Automations always run on Economy (Flex)** — about half the base price, at a lower (slower) priority. That's fixed and separate from the per-automation [Intelligence](/get-started/automate#intelligence) setting: pinning an automation to High or Low changes only its model and effort, never its priority, so it keeps the Flex discount either way. | Model | Input | Cached input | Output | | ------------- | -----: | -----------: | ------: | | gpt-5.6-sol | \$6.00 | \$0.60 | \$36.00 | | gpt-5.6-terra | \$2.40 | \$0.24 | \$14.40 | | gpt-5.6-luna | \$0.24 | \$0.024 | \$1.44 | | Model | Input | Cached input | Output | | ------------- | -----: | -----------: | ------: | | gpt-5.6-sol | \$3.00 | \$0.30 | \$18.00 | | gpt-5.6-terra | \$1.20 | \$0.12 | \$7.20 | | gpt-5.6-luna | \$0.12 | \$0.012 | \$0.72 | | Model | Input | Cached input | Output | | ------------- | ------: | -----------: | ------: | | gpt-5.6-sol | \$12.00 | \$1.20 | \$72.00 | | gpt-5.6-terra | \$4.80 | \$0.48 | \$28.80 | | gpt-5.6-luna | \$0.48 | \$0.048 | \$2.88 | For unusually large GPT-5.6 requests with more than 272,000 input tokens in total (including cached input), the whole request uses long-context rates: input and cached input are charged at 2× the table rate, and output at 1.5×. A request with exactly 272,000 input tokens still uses the regular rates above. Specialized tools use their own model rates: Deep Research uses **gpt-5.6-terra** with public-web search. Its cost is based on the input and output tokens reported for the finished research, using the Terra rates in the tables above; there is no separate per-report list price. | Tool model | Pricing basis | Notis rate | | ---------------------------------------------- | -------------------------------------------------- | -------------------------------------------------------------: | | Nano Banana Pro (`gemini-3-pro-image-preview`) | Text input / image input / reasoning / text output | \$2.40 / \$2.40 / \$14.40 / \$14.40 per 1M tokens | | Nano Banana Pro (`gemini-3-pro-image-preview`) | Image output | \$144.00 per 1M tokens | | gpt-image-2 | Text input / cached text input / text output | \$6.00 / \$1.50 / \$12.00 per 1M tokens | | gpt-image-2 | Image input / cached image input / image output | \$9.60 / \$2.40 / \$36.00 per 1M tokens | | gpt-transcribe | Transcription | \$0.0054 per minute | | gpt-4o-mini-tts | Text input / audio output | \$0.72 / \$14.40 per 1M tokens | | eleven\_v3 | Text read aloud | \$237.60 per 1M characters (about \$0.24 for 1,000 characters) | | Firecrawl | Web page read | \$0.0038 per page | | sora-2-pro | 720p video output | \$0.36 per second | Reading text aloud and reading a web page are counted in your usage like everything else. Both used to be free, so if you use them a lot you will now see them in your usage breakdown. [Premium Tools](/tools/included-tools) are counted the same way. Each one has its own price per use, shown on its page, and it comes out of the same monthly usage. ### What happens when I reach my limit? Once your monthly usage is exhausted, Notis notifies you and prompts you to either upgrade to a higher tier, activate **On-Demand Usage**, or wait for your billing cycle to reset. On teams, that means the shared team pool is exhausted. On-Demand is optional. If you leave it disabled, Notis pauses paid work once your plan cap is exhausted. If you enable it, Notis can keep working past your included credits and bill the extra usage separately. You can choose a fixed monthly On-Demand limit or use Unlimited mode. Read the full guide here: [On-Demand Usage](/account-and-settings/billing/on-demand-usage). # Privacy & Security Source: https://help.notis.ai/account-and-settings/privacy-and-security Learn more about our security and privacy practices. At Notis, we prioritise the security and privacy of your data. This document outlines the security and privacy practices, as well as the compliance standards of the platforms we use to deliver our services. ## Privacy Approach The privacy of your data is paramount for us. To ensure maintaining it we focus on the following principles: 1. We do our best to comply with GDPR standards and only work with providers that comply with GDPR and, in particular, with EU data residency requirements. 2. We do not access your data without prior consent and have strict access control to our production environment. For certain integrations, Notis uses vetted processors: * Composio to manage authentication and execute user‑initiated actions in connected apps (including Google services). OAuth is provisioned via Composio's developer app strictly to perform the actions you enable.[\[1\]](https://www.composio.dev/privacy)[\[2\]](https://docs.composio.dev/getting-started/welcome) * supermemory to provide the long‑term memory layer for personalized experiences Both process data only to deliver the features you request. You can disconnect integrations at any time in‑app and at the provider. Tokens are stored encrypted and can be revoked at any time. [Learn more in our Privacy Policy](https://notis.ai/policies/privacy) ### Who can read my conversation with Notis * Only Flo is authorized in the production databases of our suppliers. We will always reach out to you to ask for your permission to view your conversations with Notis for debugging purposes and you'll always be free to refuse. ### Can my personal data be used to train LLMs? No, your personal data is not used to train large language models (LLMs). We adhere to strict data processing principles that ensure your conversations and personal information remain private and are not incorporated into AI training datasets. Our LLM providers (primarily OpenAI) [have confirmed](https://platform.openai.com/docs/concepts) they do not use customer data shared through their API for training purposes. ### Are you GDPR compliant? Notis is committed to protecting the privacy rights of individuals in accordance with the General Data Protection Regulation (GDPR). We have implemented the following measures: * Data Processing Agreements (DPAs) with all our service providers * Mechanisms for data subject access requests * Data retention policies that limit how long we keep your information * EU data hosting when available * Transparency about data processing activities ### What’s the deletion policy for my data? Notis respects your right to have your data deleted. We have implemented a comprehensive data deletion process that ensures your information is properly removed from our systems when requested. * **User-Initiated Deletion**: You can request the deletion of your data at any time by contacting us through our live chat. * **Deletion Process**: When we receive a deletion request, we will remove your personal data from our active systems within 30 days (most often in a couple of hours). * **Backup Retention**: Data may persist in encrypted backups for up to 90 days after deletion from active systems, after which it will be automatically purged. * **Exceptions**: We may retain certain information as required by law or for legitimate business purposes, such as fraud prevention, accounting, or legal obligations. We are committed to making the deletion process as straightforward as possible while maintaining compliance with applicable regulations. ## Security Measures We take your data security very seriously. To ensure the security of your data, we focus on the following principles: * Strict access controls of our production environment. Only Flo, the founder and creator of Notis has access to them. * All our suppliers with visibility on your discussions with Notis are requested to be SOC2 or ISO certified — exception made of Telegram and Supermemory. * Backup systems with encrypted storage. ### Secure File Sharing & Link Mechanisms Notis utilizes Supabase to host and manage all attachments, audio files, and documents shared between users and Notion. Our approach to secure file sharing follows these principles: * **Unique URL Generation**: Every shared link or URL contains a unique identifier that includes the user ID and other randomized components, making it practically impossible to guess or brute force. * **Public Bucket Architecture**: We use public buckets to enable seamless sharing between channels (WhatsApp, Telegram, email) and Notion, which eliminates the need for complex authentication processes. * **Security Through Obscurity+**: While the links are technically accessible without authentication if someone has the exact URL, the complexity and uniqueness of each URL provide a high level of practical security against unauthorized access attempts. * **No Sign-up Required**: Recipients can access shared documents directly through the provided links without needing to create accounts or authenticate, improving user experience while maintaining security. This approach balances security needs with usability considerations. While we do not require authentication for accessing shared files, the mathematical improbability of guessing valid URLs makes this a secure solution for most use cases. For highly sensitive data, we recommend additional security measures such as encrypting the content before sharing or using Notion's native permission controls. ## Platform Providers Security Overview We use the following platform providers to deliver the Notis experience. Below you'll find the security and privacy practices of each of our partners. | **Platform** | Purpose | **Certifications** | **GDPR Compliance** | | ------------------------------------------- | -------------------------------------- | ---------------------------------------------------- | ------------------- | | [Notion](https://trustcenter.notion.com/) | Insert, update and find documents | ISO/IEC 27001, 27701, 27017, 27018, and SOC 2 Type 2 | Fully compliant | | [Supabase](https://security.supabase.com/) | Our main database and file storage | SOC 2 Type 2, HIPAA | Fully compliant | | [OpenAI](https://trust.openai.com/) | LLM and RAG | SOC 2 Type 2, CSA STAR Level 1, ISO 27701:2019 | Fully compliant | | [Render.com](http://Render.com) | Hosting | ISO/IEC 27001, SOC 2 Type 2 | Fully compliant | | [Langfuse](https://langfuse.com/security) | Tracing for debugging | ISO/IEC 27001:2022, SOC 2 Type II | Fully compliant | | [Twilio](https://security.twilio.com/) | WhatsApp integration | ISO/IEC 27001, 27017, 27018, and SOC 2 Type 2 | Fully compliant | | **Telegram** | Telegram integration | No known ISO certifications or SOC 2 compliance | Fully compliant | | [**Composio**](https://trust.composio.dev/) | Managed auth and integration execution | SOC2, ISO 27001:2022 | Fully compliant | | [supermemory](https://supermemory.ai) | Long-term memory layer for Notis PRO+ | Not yet certified | Fully compliant | | [Customer.io](http://Customer.io) | Automation & customer messaging | ISO 27001, SOC 2 Type II, HIPAA, GDPR, CCPA | Fully compliant | ## Privacy and Security Contact If you have any questions about our privacy and security practices or wish to report a security concern, please contact us through our live chat. # Referral Program Source: https://help.notis.ai/account-and-settings/referral-program Share your personal Notis link, give friends 10% off, and earn 30% of every sale. ## Find your referral link Every Notis account gets a personal referral link and a personal 10% discount code automatically. 1. Open the Notis portal. 2. Click your account in the bottom-left corner. 3. Choose **Refer & earn**. 4. Click **Copy link** or **Share link**. 5. You can also copy your personal discount code for friends to enter at checkout. Anyone who subscribes through your link saves 10%. You earn 30% of every sale, with no commission cap. Your discount code is a backup for places where a link is hard to use, such as a podcast, video, or in-person recommendation. Your friend can enter it at checkout to receive the same 10% saving and make sure the referral is credited to you. ## Follow your referrals On the **Refer & earn** page, click **Open partner dashboard**. Notis signs you in securely, so you do not need a separate password. The partner dashboard shows visits, referrals, commissions, and payouts. Your personal link keeps working when you share it in a message, post, newsletter, tutorial, or client recommendation. ## If your link is not ready Make sure your Notis account has a confirmed email, then click **Try again**. If the problem continues, contact us through the Help Center. # Teams Source: https://help.notis.ai/account-and-settings/teams Manage your team on Notis: invite teammates, view who installed automations in the team library, handle central billing, and configure team settings.
PRO+ Included with Pro+ and Ultra
Notis Teams allows you to collaborate with colleagues, share automations, and manage billing centrally. This guide covers everything you need to know about setting up and managing your team. Your team's included usage is pooled. If your plan includes \$150 of monthly usage per seat and you have 10 paid seats, your team gets one shared \$1,500 monthly pool. Any teammate can use it, and all team activity counts toward the same total. ## Getting Started with Teams ### Creating a Team To create a team, head to the Notis portal and navigate to **Teams** in the settings. From there, you can create a new team and start inviting members. All team members can only be on the same subscription tier. Notis for team is available in the PRO+ and ULTRA tier. ### Inviting Teammates You invite teammates by email. When you send an invite: * Enter the teammate's email (their name is optional) * The invitee receives an email with a link to join your team * Once onboarded, they'll have access to shared team resources You're only billed for a seat once the teammate **activates** their account (finishes onboarding). A pending invite that hasn't been accepted yet does not add a paid seat. **Resending Invites:** If a teammate hasn't joined yet, you can resend the invite from the Teams settings. They'll be prompted to complete onboarding. You can follow their onboarding status from the team page. You can only invite teammates that don't have an active subscription yet. Contact our support if you need to consolidate a teammate with an active subscription. ### Removing Team Members To remove someone from your team: 1. Go to **Teams** → **Members** 2. Find the member and click **Remove** 3. Their access to team resources is revoked immediately Removing a teammate does **not** automatically lower the number of seats you pay for — the freed seat stays available so you can invite someone else into it. When you want to stop paying for it, lower your subscriber count on the **Subscribe** page. You can't set fewer seats than the number of active members on your team. ## Managing Team Seats You control how many seats you pay for from the **Subscribe** page, in the **Team Seats** card, using the + / − stepper: * A seat is **added automatically** when an invited teammate activates their account. * Removing a teammate never reduces your seats automatically — use the stepper to lower them when you're ready. * The minimum is the number of active members currently on your team. * Increasing seats charges the prorated amount immediately; decreasing takes effect without an immediate refund (the seat was already paid for the current cycle). Only the team owner can change the shared plan, paid seat count, or On-Demand settings. Teammates can see the team's shared plan and usage, but should ask the owner when billing needs to change. ## Team Usage and Spend Limits The **Team** page shows how the shared pool is being used, and lets you cap any teammate's share of it. ### Seeing who is using what Under **Usage this period** you get one row per teammate: what they have spent this billing period, their share of the pool, and their limit if they have one. Teammates see only their own row. For a closer look, open the **Usage** page. As the team owner you get a scope switcher next to the date picker: * **My usage** — your own activity, exactly as before * **Entire team** — every teammate's activity together, with a **Member** column and a per-member breakdown you can click into * **Any teammate by name** — one person's activity on its own Usage views show costs, timestamps, channels and statuses — never message content. You cannot read a teammate's conversations from the Usage page, the Team page, or the API, and neither can anyone else on the team. That applies to owners too. ### Setting a spend limit In the **Limit** column, switch a teammate from **No limit** to **Fixed** and enter a dollar amount, then save. * **No limit** (the default) means they draw from the shared pool like everyone else. * A fixed limit blocks them once their own spend for the period reaches it, even if the pool still has credit. * **\$0** is a real setting: it stops them spending entirely. * Limits reset with your billing period, and you can change or remove one at any time. When a teammate hits their limit they keep full access to the portal, but new work is answered by the account assistant explaining that their limit is reached and to ask you to raise it. Their scheduled automations are skipped and voice calls are declined until you change it or the period rolls over. Limits are ceilings, not reservations. You can hand out limits that add up to more than the pool — the pool still runs out first. The Team page warns you when the total goes over so it is a deliberate choice. You can set a limit on your own account too. Notis asks you to confirm first, because past that point you get routed to the account assistant like anyone else. During a trial the whole team shares one flat pool rather than a per-seat amount, so limits above that pool never come into play. ## Team Automations Library The Team Automations Library lets your team share and discover automations created by teammates. ### Browsing the Library * Open **Automations** in the portal — templates shared by your team appear in the **Team templates** section * See all automations shared by your team, with their trigger and delivery channel * Templates you already installed show an **Installed** badge instead of the install button ### Sharing Your Automations To share an automation with your team: 1. Create or open an existing automation 2. In the automation menu, select “Publish as template” 3. You can select the name, description, and editing privilege of your template. The template creator always keeps editing access. You can let everyone on the team edit, or choose specific teammates. ### Installing Team Automations When you find an automation you want to use: 1. Click **Install** on the automation card 2. A linked automation is added to your personal automations and the editor opens 3. Review the delivery channel and schedule to fit your workflow Automations created from a template stay linked to it. Template updates to the name, instructions, intelligence, and conversation context sync automatically to every linked automation. Your own delivery channel and schedule stay personal and are not overwritten. If you have permission to edit the template, editing a linked automation also lets you update those shared fields for the whole team. If you only want to change the instructions for yourself, duplicate it into a new automation first. ## Sharing apps, skills, and automations Everything your team shares stays private to the team — people outside it cannot see or install anything. Each resource type has its own home: * **Apps** — the App Store's **Team** channel. See [Publish an app](/apps/publishing) and [Install from the Store](/apps/install-from-store). * **Skills** — the **Skills** page: publish from a skill's menu, install from the **Team skills** section. See [Skills → Share a custom skill with your team](/tools/skills#share-a-custom-skill-with-your-team). * **Automation templates** — the **Automations** page's **Team templates** section, described above. # User Settings Source: https://help.notis.ai/account-and-settings/user-settings Customize your account, beta access, and Your Computer preferences. ## Manage your Settings in the Portal CleanShot 2025-08-21 at 13.56.13@2x.png Open the Notis portal and click **Account Settings** in the bottom-left navigation area. The portal settings include: * **Email and email aliases:** this is the email recognized by Notis for [sending](https://www.notion.so/23f6f212c58c80dda445c0772ac819f9) you and [receiving emails](/channels/email). * **Full name:** the name Notis uses when addressing you. * **Custom Prompt:** global instructions Notis should follow across all tasks. Use this for account-wide preferences; for Notion-specific behavior, prefer the description of the relevant Notion database. * **Join Beta Program:** PRO+ opt in to beta features with Pro+ or Ultra. Beta features can change more often than the stable product. ## Your Computer in the Desktop App Your Computer lets Notis use your local Mac for commands, browser tasks, and app control when a task cannot be completed safely or reliably in the cloud. Your Computer is included with **Ultra**. Its row always appears in Account Settings when you use the Notis desktop app: * While access is loading, the row waits without changing your current setting. * If access cannot be checked, use **Retry**. * Without Ultra, the row shows an **Upgrade to Ultra** button. Team members are asked to contact their workspace owner. * With Ultra, the row shows the activation switch. When you activate Your Computer: 1. Notis explains what local access allows. 2. You choose whether to include Desktop Use. 3. You confirm, and the setting turns on. This extra confirmation exists because local websites, files, and automation triggers can contain prompt injection attempts. Only enable Your Computer if you are comfortable letting Notis act on your computer. After Your Computer is enabled, **Approval mode** controls when local actions on your computer require your approval. If your workspace no longer includes Ultra, Notis turns Your Computer and Desktop Use off automatically. ## Agents routing **Agents routing** has its own entry in the account menu, just above **Settings**. It appears only when the feature is available for your account. That is where you connect Codex and Claude Code — in your Notis cloud computer or on your own machine through the desktop app — and where you set the account default for the work Notis delegates. Nothing else is chosen there: routing is picked per channel, per automation, and per conversation, each with its own explicit fallback checkboxes rather than a fixed order. * [Agent routing overview](/agent-routing/overview) — the five targets and what the fallbacks do. * [Connect your agents](/agent-routing/connect-agents) — signing each target in. * [Route channels, automations, and conversations](/agent-routing/route-channels-and-automations) — where to apply a target. ## Manage your Settings with Notis Notis can also adjust the following settings for you: * **Email** * **Full Name** * **Preferred Language:** this is the language set during your onboarding. You can also set your language per database by adding an instruction in the database description. * **Timezone:** the timezone is used to run your automations and reminders at the correct time. It is normally inferred from your phone country code, but you can ask Notis to set a specific timezone. * **Dictionary:** Notis can manage a dictionary of words the speech engine does not recognize correctly. Simply ask Notis something like: "Can you remember that Notice is spelled Notis?" ## Delete Account You can permanently delete your Notis account from the portal under user settings. This action is irreversible and will remove all your data, automations, and connected integrations. ## Remove All Memories You can permanently delete all your Notis long term memory. This action is irreversible and will remove all saved memories, including the ones you uploaded from the portal. # Connect your agents Source: https://help.notis.ai/agent-routing/connect-agents Sign Codex and Claude Code in to Notis — device code in the cloud, your own CLI login on your machine, never an API key. Every target except **Notis** runs on a coding-agent account you sign in yourself. The two cloud targets are connected inside Notis; the two local ones reuse the CLI login that already exists on your machine. All four are shown in one place: open your account menu and choose **Agents routing**. Each target has its own status card and, when it is not ready, says what it needs. Notis never asks for an API key and never accepts one. Every coding-agent target runs on your own Codex or Claude subscription, so usage lands on the account you already pay for. The one credential you ever paste — the Claude setup token below — is a subscription login minted by Claude Code itself, not an API key. ## Codex Cloud Codex Cloud signs in with a device code, the same way you would sign a TV app in to a streaming service. Nothing is typed into Notis. Account menu → **Agents routing** → the **Codex Cloud** card. Press the button on the card. It reads **Connect Codex Cloud** the first time, **Continue authorization** while an authorization is still waiting on you, and **Reconnect** after an attempt expired or failed — all three start or resume the same flow. Notis prepares your cloud computer and installs the Codex CLI there, which takes a few seconds the first time. The card shows a verification link and a short code, and Notis opens that link for you. Enter the code and approve with the ChatGPT account whose Codex subscription you want to use. The copy button next to the code puts it on your clipboard. The card flips to **Connected**. The authorization code is single-use and expires after about 20 minutes — if you miss the window, press **Reconnect** and start again. Press **Disconnect** on the connected Codex Cloud card. This signs Codex out without deleting your cloud files. Press **Connect** again and approve with the other ChatGPT account. Your Codex credential is written by the Codex CLI itself, inside your own cloud computer. Notis only ever handles the verification link and the short code. ## Claude Cloud Claude Code has no device-code flow, so the credential comes from a token you mint yourself. In a terminal, run: ```bash theme={null} claude setup-token ``` Claude Code opens your browser, you log in with your normal Claude subscription, and the command prints a token. Account menu → **Agents routing** → the **Claude Cloud** card → paste only the token and press **Connect Claude Cloud**. If your terminal wrapped it onto two lines, paste both lines — Notis joins them automatically. The field is masked, and Notis never shows the token again. Notis installs the Claude Code CLI in your cloud computer, stores the token there, and runs a check. If the token is good, the card flips to **Connected** and the button becomes **Disconnect**. If Claude refuses the token, the card says so and nothing is stored. What this means in practice: * **Notis never sees your Claude login.** You log in to Anthropic in your own browser. Notis receives only the token that command printed. * **The token lives in your cloud computer, nowhere else.** It is written to a private file that only your sandbox can read. It is not stored in the Notis database, not shown again after you save it, and not written to any log. * **It lasts one year.** Anthropic issues `setup-token` credentials with a one-year lifetime, and there is no renewal notice — the first sign that it expired is a Claude Cloud run failing. Re-run `claude setup-token`, paste the new value, and you are back. If you want a reminder, ask Notis to set one for eleven months out. * **You can disconnect at any time.** Removing the connection deletes the token from your cloud computer. Anthropic documents `claude setup-token` as the supported way to authenticate "CI pipelines, scripts, or other environments where interactive browser login isn't available". That is exactly what your cloud computer is. ## Codex Local and Claude Local The local targets have nothing to connect inside Notis — they reuse the CLI you already use in your own terminal. You need three things: 1. **The CLI installed and logged in.** Run `codex login` or `claude` once in your terminal and sign in with your subscription. Confirm it works there before expecting Notis to use it. 2. **The Notis desktop app running and signed in**, with **Your Computer** enabled. That toggle is the trust boundary — see [Your Computer](/your-computer/overview) for the full setup. 3. **The desktop app awake.** A sleeping laptop makes a Local target unavailable. Pin Auto if you want Notis to keep answering elsewhere; pin the Local target if you want the request to fail rather than move. Notis never reads, copies, or transmits your local CLI credential. It asks the CLI to run a task and reads back the result. ## Reading the status cards Each card carries a status badge, and often a line of detail under it explaining the badge. | Status | Where it appears | What it means | | ---------------------------------------- | ---------------- | ------------------------------------------------------------------------------------------------------------------------------------------- | | **Connected** | Any target | Ready to be routed to. | | **Not connected** | Cloud targets | Nothing has been connected yet, or the credential was removed. Connect it here. | | **Sign in needed** | Local targets | The CLI is installed on your machine but not signed in. Run `codex login` or `claude` in your terminal. | | **Not installed** | Any target | The CLI is missing — install Codex or Claude Code locally, or press the connect button again so Notis reinstalls it in your cloud computer. | | **Disconnected** | Local targets | The Notis desktop app is not reachable — usually asleep, quit, or **Your Computer** turned off. | | **Update required** | Local targets | The desktop app is too old to run delegated tasks. Update it. | | **Starting** / **Authorization pending** | Codex Cloud | The connection is in progress. **Authorization pending** means the link and code are on the card, waiting for you. | | **Expired** | Codex Cloud | The authorization was not approved inside its 20-minute window. Press **Reconnect**. | | **Needs attention** | Codex Cloud | The authorization did not complete. Press **Reconnect**. | | **Unavailable** | Any target | Notis could not read the status just now. Press **Check again**. | Any other badge is passed straight through from the runtime check; the detail line under it says what happened. A target that is not ready still shows up in every routing picker, dimmed and marked **NOT CONNECTED**. Clicking it opens **Agents routing** instead of selecting it — or, if you are already there, points you at the cards above — so you cannot accidentally aim a channel at something that cannot run. ## Related * [Agent routing overview](/agent-routing/overview) — Auto versus the five named targets. * [Route channels, automations, and conversations](/agent-routing/route-channels-and-automations) — where to apply a target. * [Your Computer](/your-computer/overview) — enabling the desktop bridge. * [Cloud Computer](/tools/cloud-computer) — the container the cloud targets run in. # Overview Source: https://help.notis.ai/agent-routing/overview Choose which agent runs your work — Auto, Notis, Codex, or Claude Code — and whether it runs on your computer or in your Notis cloud. When you give Notis a real piece of work — refactor a repo, dig through a project, run a long research task — Notis hands it to an agent. Agent routing is where you say *which* agent that is, and *where* it runs. There are six choices: | Target | What runs | Where it runs | | ---------------- | ------------------------ | ----------------------------------------------- | | **Auto** | Notis picks | Local before cloud, then hosted Notis | | **Notis** | The hosted Notis agent | Notis servers | | **Codex Cloud** | Your Codex account | Your Notis cloud computer | | **Claude Cloud** | Your Claude Code account | Your Notis cloud computer | | **Codex Local** | Your Codex CLI | Your own machine, through the Notis desktop app | | **Claude Local** | Your Claude Code CLI | Your own machine, through the Notis desktop app | ## Auto versus a named target **Auto** is the default. Notis ranks every coding-agent target your account can actually reach — your machine first, then your cloud computer, preferring whichever subscription still has usage left — and walks that chain. If nothing is connected, Auto lands on hosted Notis. Auto never fails: there is always somewhere to run. A **named target** is a promise. If you pick Codex Local and your laptop is asleep, the request fails and says so. It does not quietly answer from a machine that has never seen your repository. That is deliberate: the old behaviour silently walked Desktop → Cloud → Notis, which could move a local coding task onto a host with none of your files. There are no fallback checkboxes. Choosing Auto *is* the fallback ladder; choosing a named target is the hard pin. ## Local versus cloud **Local** targets run the coding agent you already installed and signed in to, on your own computer. Notis sends it the task and reads back the result — your Codex or Claude login stays in that CLI's own credential store and never leaves the machine. The agent gets whatever your local CLI allows: your repositories, your files, your tools. It needs the [Notis desktop app](/your-computer/overview) running. **Cloud** targets run the same first-party CLI inside your [Cloud Computer](/tools/cloud-computer) — an isolated container that belongs to your account. It stays available when your laptop is asleep, and the agent is boxed inside the container rather than sitting on your machine. Cloud targets need a plan that includes the Cloud Computer. Either way, the model bill goes to your own Codex or Claude subscription. Notis never asks for an API key and never accepts one. **Notis** is the hosted agent: no coding-agent account required. It is also where Auto ends when nothing else is ready. ## Where routing is decided Notis resolves routing for each turn in this order, first match wins: 1. **This conversation** — the agent picker in the Notis overlay or chat sidepanel. 2. **This channel or automation** — the default you set on a channel, or on a single automation. 3. **Account default** — usually Auto after migration; Agents routing no longer sets this from the UI. So a channel pinned to Codex Local with Auto everywhere else is a normal setup, and picking Claude Cloud once from the overlay does not change anything permanently. ## Related * [Connect your agents](/agent-routing/connect-agents) — sign each target in. * [Route channels, automations, and conversations](/agent-routing/route-channels-and-automations) — set routing per scope. * [Your Computer](/your-computer/overview) — what the desktop app makes reachable. * [Cloud Computer](/tools/cloud-computer) — the isolated container the cloud targets use. # Route channels, automations, and conversations Source: https://help.notis.ai/agent-routing/route-channels-and-automations Set Auto or a named agent for one channel, one automation, or a single conversation. Routing is set in three places, from broadest to narrowest. Each one falls back to the one above it, so you only set what you actually want to differ. | Scope | Where you set it | | ---------------- | ------------------------------------------------- | | One channel | Channels page → the channel's defaults | | One automation | The automation editor | | One conversation | The agent picker in the overlay or chat sidepanel | At request time Notis takes the narrowest choice that exists: conversation, then channel or automation, then the account default (Auto unless you have changed it elsewhere). **Agents routing** in the account menu is only for connecting Codex and Claude. It does not pick an account-wide target — that default is Auto. ## One channel Open the **Channels** page, then the defaults for the channel you want — WhatsApp, Telegram, Slack, email, the desktop app, and so on. The **Agent** section lists Auto, then Notis, Codex Cloud, Claude Cloud, Codex Local, and Claude Local. Leave it on **Auto** unless you want that channel pinned. Save, and every message arriving on that channel routes there. This is the useful one for splitting work by context: WhatsApp on Auto for everyday chat, the desktop app on **Codex Local** because that is where your repositories are. ## One automation Open the automation and set its **Agent** in the editor. * An automation that must touch your local machine — a nightly repository job, something reading local files — should pin a Local target. If the machine is offline, the run fails loudly instead of answering from a host without your files. * An automation that just needs *an* answer — a daily digest, a summary — should stay on **Auto** so a sleeping laptop does not skip your morning brief. An automation with no routing set inherits Auto (or whatever narrower scope sits above it). ## One conversation In the Notis overlay and in the chat sidepanel, the composer has an agent picker alongside the intelligence and speed pickers. Open it with `⌥ A`, or click it, then choose with the arrow keys or the number next to each target. Two things to know: * **It only affects that conversation.** Nothing is saved to the channel or your account. * **Unconnected targets are dimmed.** Clicking one takes you to **Agents routing** to connect it, rather than selecting a target that cannot run. The button next to the composer names the current target whenever it is not plain Auto/Notis inheritance, so you can always see where the next message is going. A dashed outline means the target is inherited from your channel; a solid one means you picked it for this conversation. ## A worked example A setup that keeps local work local without making everything brittle: * **Everywhere else:** Auto. Everyday chat always answers, and prefers your machine when it is awake. * **Desktop app channel:** Codex Local. Coding work runs on your machine, or fails rather than being answered by an agent with no access to your code. * **Nightly repository automation:** Codex Local. It runs where the repository is, or it tells you it did not run. ## Related * [Agent routing overview](/agent-routing/overview) — Auto versus named targets. * [Connect your agents](/agent-routing/connect-agents) — signing each target in. * [Automations](/get-started/automate) — building the automations you route. * [Channels compared](/channels/channels-compared) — the channels you can set defaults on. # Ship Apps Source: https://help.notis.ai/agents/build-and-ship-apps Use your favorite coding agent to build apps and deploy them to Notis Cloud. Coding agents are great at writing app code. Notis gives that code somewhere useful to live. Building and deploying a Notis App from your coding agent runs through the Notis CLI and is included with Pro+ and Ultra. New accounts can try it for their first 7 days before picking a plan. Build with ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, or another local agent, then deploy the app to Notis Cloud and share it with your team. ## What this solves When you build an app from scratch, you still have to figure out hosting, databases, auth, permissions, deployment, and team distribution. Notis Apps packages those pieces into one installable product inside Notis. ## What your agent can build A Notis App can include: * A custom UI inside the Notis sidebar. * Databases for the app's data. * Automations and skills. * Integration access through Notis. * Listing details for sharing or publishing. ## Start from Notis 1. Open the [Notis portal](https://app.notis.ai). 2. Click **+ Create app**. 3. Choose **your local code agent**. 4. Copy the generated prompt into ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, or your agent of choice. The prompt tells the agent how to use the Notis CLI to create, develop, and deploy the app. ## Ship it Your agent builds and deploys the app to your Notis workspace. From there, you can: * Use it yourself. * Share it with your team. * Publish it in the team store. * Submit it to the public Store after reviewing the final Store page. ## Related * [Notis Apps](/apps/overview) * [With Notis](/apps/build-with-notis) * [Publish an app](/apps/publishing) # Sync Integrations Source: https://help.notis.ai/agents/connect-integrations Connect apps once in Notis, then use them from ChatGPT, Claude, Grok, Notion, or your coding agent. Notis gives your coding agents one shared path to the apps and databases connected to your account. Instead of reconnecting Gmail, Slack, GitHub, Notion, Linear, Calendar, and MCP servers separately for every agent, connect each service once in Notis. ## Choose how your agent connects * **Notis CLI** — Best for ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, and other agents that can run commands. * **Remote MCP** — Best for ChatGPT on the web, Claude, Grok, Notion, and other apps with an MCP connections page. Both use the same account. You can use the CLI in one agent and Remote MCP in another without connecting your apps again. ## Start with the Notis CLI Give this command to your coding agent: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis start ``` Your agent asks for your email, guides you through sign-in, and gives you a secure Notis link for each app you want to connect. The sign-in link opens the [Notis desktop app](/channels/manager), which is what signs the Notis CLI in on your machine. Install it first if you do not have it yet. See [Start with the Notis CLI or Remote MCP](/agents/overview) for the full walkthrough. ## Use it with an agent For example: ```text theme={null} Use the Notis CLI to find my latest GitHub issues and draft a Slack update. ``` The agent finds the right action and runs it through your Notis account. You do not need to configure each app again inside every agent. ## Connect a new integration If an integration is not connected yet, your agent gives you a Notis link. Open it, approve the connection, and ask the agent to retry. You can also connect and manage integrations from the Notis portal at **Settings → Integrations**. See [Integrations](/get-started/integrations) for the full help-center guide, including multiple accounts and labels. Once connected, the integration is available through both the Notis CLI and Remote MCP. ## Check your connection If the CLI stops working, ask your agent to run: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis doctor ``` If the check says you are signed out, open the [Notis desktop app](/channels/manager) and sign in again. If you have not installed it yet, install it first: it is what signs the Notis CLI in on your machine. ## Related * [Start with the Notis CLI or Remote MCP](/agents/overview) * [Keep skills in sync](/agents/keep-skills-in-sync) * [Ship Apps](/agents/build-and-ship-apps) # Sync Skills Source: https://help.notis.ai/agents/keep-skills-in-sync Keep the right Notis skills available in Codex, Claude Code, Cursor, and Notis. Skills Sync keeps your agent instructions in one place. Create or install a skill once, then choose which coding agents should have it. This saves you from copying skills between agents, losing track of the latest version, or realizing mid-task that an agent does not know the workflow you already wrote somewhere else. Creating, installing, and syncing skills — including pushing a local skill back up to Notis — is included with Pro+ and Ultra. A skill tied to an Ultra feature also needs Ultra; for example, Desktop Use only syncs when Your Computer is available. ## How it helps * **One shared skill library** — Skills you create in Notis can appear in Notis, Codex, Claude Code, and Cursor. * **Local skills come back to Notis** — Skills you write for a local agent can sync back to your Notis account. * **Choose the right agents** — Send each skill only to the agents that need it. * **No duplicate cleanup** — Notis keeps one canonical copy when the same skill appears in multiple places. ## Turn it on Skills Sync lives on the **Skills** page in the [Notis portal](https://app.notis.ai/skills). 1. Open [Notis](https://app.notis.ai) and go to **Skills**. 2. Find the **Desktop skill sync** card. 3. Turn on **Enable sync**. 4. Click **Sync Now**, or open the Notis desktop app and let it sync automatically. Enable sync toggle on the Skills page The card shows the last sync time after the first run. ## What syncs * Skills installed or created in Notis. * Local skills from Notis, Codex, Claude Code, and Cursor. * Built-in skills such as `notis-cli` and `notis-apps`, when they are available to your account. Curated skills from the Notis catalog are linked locally for use, but edited in the portal. ## Pick which agents get a skill Open a skill in the [Notis portal](https://app.notis.ai/skills) and choose its agent targets: * Notis * Codex * Claude Code * Cursor Turning an agent off removes the local link for that agent. It does not delete the skill. ## Pause sync Go back to the **Desktop skill sync** card and turn **Enable sync** off. Existing local skills stay where they are, but new changes stop syncing until you turn it back on. ## FAQs Delete skills from the **Skills** page in the [Notis portal](https://app.notis.ai/skills). Deleting a skill directly inside Codex, Claude Code, or Cursor usually will not stick. On the next sync, Notis recreates the local link from your Notis account. No for skills you own. Local edits are detected and pushed back to your Notis account on the next sync. Curated catalog skills are read-only locally. Edit those from the portal instead. It is removed from your local agent folders on the next sync. Notis keeps the local mirror as a safety net, so files are not immediately wiped from disk. Yes. Open the skill in Notis and turn off that agent's target. This removes the local link for that agent without deleting the skill from Notis or from your other agents. Yes. Sign in to the Notis desktop app on each computer with the same account. Each computer keeps its own local mirror, and changes sync through your Notis account. Notis only manages links it owns. If a real folder or non-Notis link already exists at that path, Notis skips it and continues syncing the rest. Move the conflicting folder if you want Notis to manage that skill slot. ## Related * [Sync Integrations](/agents/connect-integrations) * [Ship Apps](/agents/build-and-ship-apps) # Start Source: https://help.notis.ai/agents/overview Connect Notis to ChatGPT, Claude, Grok, Notion, or a coding agent with the Notis CLI or Remote MCP. Switch coding agents without rebuilding your setup. Notis gives each agent the same integrations, skills, and long-term memory through your Notis account. You can connect in two ways: * **Notis CLI** — Use this with ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, or another agent that can run commands on your computer. * **Remote MCP** — Use this with ChatGPT on the web, Claude, Grok, Notion, or another app that lets you add an MCP connection. Both use the same Notis account. Choose the connection your agent supports; your tools and context stay the same. ## Connect with the Notis CLI Start a session in ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, or OpenCode. The session must be allowed to run commands on your computer. ```bash theme={null} npx --package @notis_ai/cli@latest -- notis start ``` Paste the command into the conversation and ask the agent to run it. If your ChatGPT desktop session cannot run commands, use the ChatGPT Remote MCP steps below instead. Your agent asks for your email address. Notis sends you a secure sign-in link instead of making you complete a separate signup form. The sign-in link opens the [Notis desktop app](/channels/manager), which is what signs the Notis CLI in on this machine. Install it first if you do not have it yet: the CLI waits for that sign-in before it continues. Your agent gives you a Notis link for each app. Open it, approve the connection, then return to the agent. Ask the agent to use Notis whenever it needs Gmail, Calendar, Notion, Slack, GitHub, Linear, or another connected service. The same command works in every supported coding agent: * **ChatGPT desktop** — Open a coding session with terminal access, paste the command, and ask ChatGPT to run it. * **Claude Code** — Open Claude Code in your project, paste the command, and follow its prompts. * **Cursor** — Open your project in Cursor, paste the command into Agent chat, and ask it to run the command. * **OpenClaw** — Start an OpenClaw session that can use your terminal, then paste the command. * **Hermes** — Start Hermes with command access, then paste the command. * **OpenCode** — Open your project in OpenCode, paste the command, and follow its prompts. ## Connect with Remote MCP Use this address for every Remote MCP connection: ```text theme={null} https://mcp.notis.ai/mcp ``` After you add it, your browser opens Notis. Sign in, check the agent name and permissions, then choose **Allow access**. ### ChatGPT on the web Full Notis actions require ChatGPT Business, Enterprise, or Edu. A workspace admin or owner must allow developer mode first. 1. In ChatGPT, open **Settings → Apps → Advanced Settings**. 2. Turn on **Developer mode**. If the setting is missing, ask your workspace admin to allow it. 3. Return to **Settings → Apps** and choose **Create**. 4. Name the app **Notis** and paste the Notis MCP address. 5. Create the app, review the tools, and sign in to Notis. ### Claude On Pro or Max: 1. Open **Settings → Connectors**. 2. Choose **Add custom connector**. 3. Paste the Notis MCP address and choose **Add**. 4. Connect and sign in to Notis. On Team or Enterprise, an owner opens **Settings → Connectors**, switches to **Organization connectors**, and adds Notis. Members can then find Notis in **Settings → Connectors** and choose **Connect**. ### Grok 1. Open [grok.com/connectors](https://grok.com/connectors). 2. Choose **New Connector → Custom**. 3. Paste the Notis MCP address. 4. Finish the Notis sign-in. On Grok Business or Enterprise, a team admin must make the connector available to the organization first. ### Notion Notion MCP connections require a Business or Enterprise plan. Your workspace must also allow custom MCP connections. 1. Open **Settings → Connections → Discover**. 2. Choose **Add Custom MCP**. 3. Enter **Notis** and paste the Notis MCP address. 4. Finish the Notis sign-in. If **Add Custom MCP** is missing, ask your workspace admin to allow custom connections. ### Another MCP app Open the app's connections or tools settings, add a remote MCP server named **Notis**, and paste the Notis MCP address. ## What happens during Remote MCP setup Paste the Notis MCP address into your app and name the connection **Notis** if it asks for a name. Your browser opens Notis. Sign in if needed. Check the agent name, the Notis account being connected, and each permission it requests. Choose **Allow access** only if you recognise the agent. The agent finishes the connection and can start finding the right Notis tools for your requests. ## What the Remote MCP permissions mean * **Read through Notis** — Find and read information in Notis and services you connected to Notis. * **Take actions through Notis** — Create or change information when you ask the agent to take an action. * **Help connect services** — Give you a Notis connection link when a service still needs your approval. The agent only receives the permissions shown on the approval screen. If it later needs another permission, Notis asks you to approve the connection again. ## When a service is not connected If the agent needs Gmail, Notion, Slack, Calendar, or another service that is not connected yet, it gives you a Notis link. Open it, approve the service, then ask the agent to try again. Never paste a password or private access key into the agent. ## What follows you between agents * **1,000+ integrations** — Connect an app once, then use it through the CLI or Remote MCP. * **Skills** — Create or install a skill once and keep it available in Notis, Claude Code, Codex, and Cursor. Creating, installing, and syncing skills through the desktop app is included with Pro+ and Ultra. Skills for Ultra features, such as Desktop Use, also need Ultra. * **Long-term memory** — Give each agent the useful context Notis has learned instead of explaining it again. * **Notis Apps** — Build with your favorite coding agent and deploy the result into Notis, through the Notis CLI. Included with Pro+ and Ultra. ## When you need the desktop app Remote MCP does not need it: you sign in to Notis in your browser and never install anything. The Notis CLI does. The [Notis desktop app](/channels/manager) is what signs the CLI in on your machine and keeps that session fresh, so install it before running `notis start`. It also syncs skills into local agent folders, builds Notis Apps locally, and connects tools that run on your computer. ## Remove or reconnect an agent To remove Remote MCP access: 1. Open Notis. 2. Open your account menu. 3. Choose **Notis CLI/MCP**. 4. Find the agent and choose **Remove access**. 5. Confirm the removal. The agent loses its Notis access immediately. Your services stay connected to Notis, so your other Notis features continue to work. To reconnect or change permissions, remove the old connection and add Notis again from the agent's settings. Notis shows a fresh approval screen. ## Troubleshooting * **The CLI is not working** — Ask your agent to run `npx --package @notis_ai/cli@latest -- notis doctor`. * **The Remote MCP approval expired** — Return to the agent and start the connection again. * **The wrong Notis account appears** — Cancel, sign out of Notis in that browser, then restart the connection. * **The agent says a permission is missing** — Reconnect Remote MCP and approve the permission you want it to use. * **A service is unavailable** — Open the connection link from the agent, finish connecting the service, then retry. ## Related * [Sync Integrations](/agents/connect-integrations) * [Keep skills in sync](/agents/keep-skills-in-sync) * [Build and ship Notis Apps](/agents/build-and-ship-apps) # With Notis Source: https://help.notis.ai/apps/build-with-notis Describe what you want, and let Notis build the app for you in a sandbox. This is the no-setup path. You describe the app you want in chat, and Notis writes the code, builds it, deploys it, and installs it in your workspace. You don't need a terminal or a local agent. Building and editing apps with Notis is included with PRO+ and ULTRA. PRO users can install and use Store apps, then upgrade from the setup chat when they want to customize one. ## Start the wizard 1. Open the [Notis portal](https://app.notis.ai) and click **+ Create app** in the App Store (or the `+` shortcut next to *Store* in the sidebar's Customize section). 2. Pick a **name** and an optional **starting point**: * **Empty scaffold** — a bare app you fill in from scratch. * **A scaffold from the catalog** — Databases, 5 Minutes Journal, Notis Notes, or Random Number Generator. 3. Choose **Notis** as the builder. 4. Click **Open Notis chat** to send your prompt to the assistant. The wizard prompts you to describe what you want, e.g. *"Build me a CRM app with companies, contacts, and deals, and a dashboard that shows my top opportunities."* ## What happens behind the scenes 1. Notis triages your request and routes it to the **app-building agent**. 2. A **sandbox session** is created for you (one persistent sandbox per user, reused across conversation turns). 3. The agent runs the Notis CLI inside the sandbox to scaffold the app, pick a starting point, fill in the config, build, verify, and deploy. 4. Your app appears in the sidebar, ready to use. You see the build happening in chat, with progress messages and a link to the live app the moment it's ready. ## Iterate conversationally The same sandbox is reused across turns, so you can refine without losing context. Try things like: * *"Add a priority field to the tasks database."* * *"Change the dashboard layout to show a table instead of cards."* * *"Add a route for analytics with a bar chart of deals per stage."* * *"Fix the bug where empty contacts crash the detail page."* * *"Refresh the Store screenshots and make the app icon feel more distinctive."* Each request triggers another build + deploy. The app reloads in place. When you ask Notis to prepare the app for the Store, it also creates the listing screenshots and release history that appear in App Details. You still review that Store page and choose when to publish. ## Switch between apps in the same conversation You can keep working on different apps in the same chat. Notis remembers which app you're editing and switches context automatically when you mention another one by name. ## When to switch to a local agent Building with Notis is the fastest way to get from idea to installed app. Switch to the local-agent path when: * You want full control over the code, file structure, or testing. * You're integrating with a private library or workflow that lives only on your machine. * You prefer to use your own coding assistant (Cursor, Claude Code, Codex, etc.) and your own editor. See **[Build with your local code agent](/apps/build-with-your-agent)** for the developer path. ## Publishing Once you're happy with your app, head to **App Details** (`/apps/[appId]`) to **Publish** it — to your team, or to the public App Store. See **[Publish an app](/apps/publishing)**. ## Related * [Overview](/apps/overview) — what apps are. * [With your coding agent](/apps/build-with-your-agent) — the developer path. * [Publish](/apps/publishing) — ship to team or public. * [Ship Apps](/agents/build-and-ship-apps) — the local coding-agent path. # With your coding agent Source: https://help.notis.ai/apps/build-with-your-agent Use ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, or any terminal to build Notis apps locally with the Notis CLI.
PRO+ Included with Pro+ and Ultra
This is the developer path. You work in your own editor, your own repo, with your own coding assistant. Notis gives the agent one path to create, preview, and deploy the app. ## Prerequisites * The [Notis desktop app](/channels/manager), installed and signed in. It keeps the Notis CLI authenticated. * A local code agent (ChatGPT desktop, Claude Code, Cursor, OpenClaw, Hermes, OpenCode, or just a terminal). * Node.js 18+. The desktop app also syncs the built-in `notis-cli` and `notis-apps` skills, so local agents can discover the CLI workflow and the app platform rules without extra setup. Verify the CLI is ready: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis doctor ``` If the doctor check passes, you're set. If not, open the [Notis desktop app](/channels/manager) and sign in. ## Start the wizard (or skip it) The fastest way to start is from the portal: 1. Open the [Notis portal](https://app.notis.ai) and click **+ Create app**. 2. Pick a **name** and an optional **starting point** (empty scaffold, or Databases, 5 Minutes Journal, Notis Notes, or Random Number Generator). 3. Choose **your local code agent** as the builder. 4. Click **Get the prompt** — Notis generates a copy-paste prompt tuned to your starting point. Paste it into your coding agent and let it drive the CLI for you. You can also skip the wizard entirely and drive the CLI by hand. ## Manual workflow ### 1. Scaffold a new app ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps scaffolds list npx --package @notis_ai/cli@latest -- notis apps init "My Task Manager" --from notis-database cd "my-task-manager" ``` This creates a Vite + React project with: * `notis.config.ts` — the app declaration (metadata, databases, routes, tools). * `CHANGELOG.md` — the release history shown on App Details and in the Store. * `app/` — your React pages. * `components/` — scaffolded shadcn UI components. * `vite.config.ts` — wrapped with `notisViteConfig()`. * Tailwind and `@notis/sdk` pre-configured. ### 2. Configure the app Edit `notis.config.ts`: ```typescript theme={null} import { defineNotisApp } from '@notis/sdk/config'; export default defineNotisApp({ name: 'task-manager', title: 'Task Manager', description: 'Track and manage team tasks', icon: 'phosphor:check-square', categories: ['Productivity'], tagline: 'Plan, assign, and ship team tasks.', screenshots: [ { path: 'metadata/screenshot-1.png', alt: 'Task dashboard with priorities and owners.', route: 'dashboard' }, { path: 'metadata/screenshot-2.png', alt: 'All tasks grouped by status.', route: 'tasks' }, { path: 'metadata/screenshot-3.png', alt: 'A task detail view with its activity.', route: 'tasks' }, ], databases: ['tasks'], routes: [ { path: '/', slug: 'dashboard', name: 'Dashboard', icon: 'phosphor:squares-four', default: true }, { path: '/tasks', slug: 'tasks', name: 'All Tasks', icon: 'phosphor:list', collection: { database: 'tasks', titleProperty: 'title' }, }, ], tools: ['LOCAL_NOTIS_DATABASE_QUERY'], }); ``` Also add a release version to `package.json`: ```json theme={null} { "notisAppVersion": "0.1.0" } ``` Keep release copy in `CHANGELOG.md`, newest first: ```md theme={null} ## [Initial Release] - {PR_MERGE_DATE} - Added the first task dashboard and task list. ``` ### 3. Build your pages Pages live under `app/`. Use the SDK hooks instead of calling the Notis backend yourself. For database-backed apps, `useDocuments`, `useDocument`, `useDatabaseSchema`, and `DocumentEditor` cover the common list, detail, schema, and editing flows. `Markdown` renders saved document content with the same styles as the Portal. ```tsx theme={null} 'use client'; import { useDocuments } from '@notis/sdk'; import { Card } from '@/components/ui/card'; export default function Dashboard() { const { documents, loading } = useDocuments('tasks', { pageSize: 25 }); return (

Dashboard

{loading ?

Loading…

: null} {documents.map(document => ( {document.title} ))}
); } ``` ### 4. Develop locally ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps dev ``` This opens the Notis desktop app directly to your local app when possible, and every active local app appears in the **Local development** sidebar group. Edits hot-reload. When the app is visible in **Local development**, the CLI reports it as mounted. If Notis was closed when you started, open the desktop app; the development session keeps waiting and connects when Notis is ready. ### 5. Build, capture, verify, and deploy ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps build npx --package @notis_ai/cli@latest -- notis apps screenshot npx --package @notis_ai/cli@latest -- notis apps verify npx --package @notis_ai/cli@latest -- notis apps create "Task Manager" . npx --package @notis_ai/cli@latest -- notis apps deploy ``` Screenshot declarations decide which route, sample scenario, focus area, and light or dark theme to capture. The command creates the Store-ready images for you. Use `--mode live` when the listing should show your real signed-in data rather than sample data, and review the images before publishing. Or, if the remote app already exists: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps link npx --package @notis_ai/cli@latest -- notis apps deploy ``` ### 6. Diagnose a project ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps doctor ``` Reports anything the platform expects but didn't find — missing routes, broken config, unknown tools, incomplete Store details, missing release history, or invalid screenshots. ## Forking an existing app You can start from an app you already have installed: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps pull ./ cd ./ npm install npx --package @notis_ai/cli@latest -- notis apps dev # edit, then build the saved source snapshot npx --package @notis_ai/cli@latest -- notis apps build npx --package @notis_ai/cli@latest -- notis apps create "My Fork" . npx --package @notis_ai/cli@latest -- notis apps deploy ``` Forked apps live as separate listings — existing installs of the original keep their upstream link. Apps deployed with older tooling may not have a saved source snapshot yet. Redeploy them once with the current CLI before pulling. ## Publishing When the app is ready, review **App Details** (`/apps/[appId]`) and its Store checklist. Publish there, or explicitly approve the preview and let your coding agent submit the exact deployed version: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps publish --confirm-ready ``` > Deploy and Store publication are separate. Deploy updates your installed app. Publish shares that deployed version with your team or sends it for public review. Store details live in `notis.config.ts`, `metadata/`, `CHANGELOG.md`, and `package.json`, so the preview always comes from reviewable source files. ## Related * [Sync Integrations](/agents/connect-integrations) — let local agents use your Notis integrations. * [Overview](/apps/overview) — what apps are. * [With Notis](/apps/build-with-notis) — the chat path. * [Publish](/apps/publishing) — ship to team or public. # Install Source: https://help.notis.ai/apps/install-from-store Browse, install, and manage Notis apps from the App Store. The **App Store** is where you discover Notis apps — public ones built by the community, or private ones shared inside your team. Installing on PRO sets up the app's pages and databases. On PRO+, it can also turn on bundled automations and skills after your approval. ## Browse the store Open the **App Store** from the sidebar (Customize → *Store*) or go to `/store` in the [Notis portal](https://app.notis.ai). You can: * Filter by category (Productivity, Sales & Marketing, Operations, Product & Engineering, or Personal). * Switch between **Public** apps and apps shared with your **Team**. * Open any listing to see its screenshots, publisher, What's New, Version History, ratings, reviews, and the pages, databases, automations, and skills it includes. * Use **Share** on a listing when you want to send that Store page to someone else. ## Install an app 1. Open a listing. 2. Click **Install**. 3. Review the resources the app will create in your workspace (databases, automations, skills) and approve. What happens next: * A new app record is created in your workspace. * The app's databases are **cloned** into your workspace — your data stays yours; the app starts from clean templates. * On PRO+, bundled automations and skills are installed. On PRO, Notis shows what is available and offers an upgrade when setup needs one of them. * The app appears in your sidebar under *Apps* with its routes ready to use. * If the app offers guided setup, its app menu includes a setup action. For example, **Set up Journal** opens Notis with the right instructions already attached. You can also ask Notis directly: * *"Show me available apps."* * *"Install the project tracker app."* ## Update an installed app When the app's publisher ships a new version, Notis brings clean installs up to date automatically. Open **App Details** (`/apps/[appId]`) to see the current version. * **Clean update** — the latest published version is applied automatically while your own data stays in your workspace. * **Resolve with Notis** — if your customizations conflict with the update, App Details flags it and Notis walks you through what changed so you can keep your edits while pulling in upstream improvements. ## Reset to the published version If your customizations have drifted and you want to start fresh, open **App Details → ⋯ menu → Reset**. The installed app returns to the latest published version while your database entries and documents stay in place. By default, Notis also preserves database fields you added yourself. Choose **Restore database schema** only when you want the app's latest published database structure too. ## Rate an app Open a published Store listing and choose **Rate This App**. Pick one to five stars and, if you want, add a short title and review. You can edit your rating later. Each person has one rating per listing, so editing replaces your earlier rating rather than adding a duplicate. ## Uninstall an app From the sidebar, right-click the app and pick **Uninstall** — or use the **⋯ menu** on the App Details page. What gets removed: * The app record and its sidebar routes. * The app's automations and skills. What stays: * The **databases** cloned into your workspace — they're yours now. If you don't need them, delete them manually from the database list. ## Related * [Notis Apps](/apps/overview) — what apps are. * [Build with Notis](/apps/build-with-notis) — create your own app by chatting. * [Build with your local code agent](/apps/build-with-your-agent) — create your own app in code. * [Publish an app](/apps/publishing) — ship your app to the store. # Overview Source: https://help.notis.ai/apps/overview Installable products that bundle a UI, databases, automations, and skills into one packaged experience inside Notis. A Notis App is the top-level packaging unit inside Notis. Think of it as a mini-product that lives in your workspace: it has a name, an icon, its own pages in the sidebar, its own databases, and — when you want it to — its own automations and skills. You can: * **Install** apps from the App Store with one click. * **Build** your own apps, either by chatting with Notis or by coding them with your favorite local agent. * **Publish** what you built to your team or to the public App Store. You can browse the Store on any signed-in plan. PRO lets you install and use apps. PRO+ adds building and editing with Notis, database customization, skills, automations, and publishing — including developing and deploying apps from Claude Code, Codex, Cursor, or another coding agent with the Notis CLI. New accounts can try the builder for their first 7 days before picking a plan. ## What an app contains An app is a single bundle that can include: * **A UI** — a small Vite + React app that renders inside the Notis Portal sidebar. * **Databases** — first-class data stores the app owns (tasks, notes, contacts, deals…). * **Routes** — the pages and sub-views you see in the sidebar. * **Automations** — workflows the app sets up for you when installed. * **Skills** — instructions and prompts the app brings along. * **Listing details** — an icon, screenshots, tagline, categories, What's New, and full Version History. * **Setup** — an optional guided conversation that helps you configure the app after installation. Apps are versioned: every deploy bumps the version, and the App Store always serves a specific published version. ## Where apps live * **App Store** (`/store`) — browse public apps or apps shared with your team, read reviews, and inspect every included resource before installing. * **Your sidebar** — installed apps appear under *Apps* in the sidebar, with their own routes and sub-pages. * **App Details** (`/apps/[appId]`) — icon, screenshots, release history, included resources, visibility, and publish controls. ## Visibility Every app you create has a visibility setting on its App Details page: | Visibility | Who can install it | | ------------ | ----------------------------------------------------------- | | **Personal** | Only you. | | **Team** | Members of your Notis team. | | **Public** | Anyone with a Notis account, after the listing is approved. | You can change visibility at any time from **App Details → Visibility**. If an app is already published or waiting for review, unpublish or withdraw it before changing between Team and Public. ## Two ways to build an app There's one entry point — the **+ Create app** button in the App Store or the `+` shortcut next to *Store* in the sidebar's Customize section — and two paths from there: 1. **[Build with Notis](/apps/build-with-notis)** — describe what you want in chat, and Notis builds it for you in a sandbox. 2. **[Build with your local code agent](/apps/build-with-your-agent)** — you, Cursor, Claude Code, Codex, or another agent write the code locally. Both paths produce the exact same artifact and install into your workspace the same way. ## Installing apps from the store See **[Install apps from the store](/apps/install-from-store)** for browsing, installing, resetting, and uninstalling. ## Publishing apps you built See **[Publish an app](/apps/publishing)** for the publish-to-team and publish-to-public flows. ## Related * [Ship Apps](/agents/build-and-ship-apps) — use a coding agent to build and deploy apps to Notis Cloud. * [With Notis](/apps/build-with-notis) — the no-code/chat path. * [With your coding agent](/apps/build-with-your-agent) — the developer path. * [Publish](/apps/publishing) — ship your app to team or public. # Publish Source: https://help.notis.ai/apps/publishing Prepare a complete Store page, then share your Notis app with your team or submit it to the public App Store. Publishing turns an app in your workspace into something other people can install. Team apps go live for your teammates immediately. Public apps go through review first. ## Prepare the Store page Open **App Details** after your latest deployment. The Store checklist shows exactly what is ready and what still needs attention: * A title, description, tagline, and at least one category. * Three to six screenshots. Each screenshot needs a useful description and must use the Store's standard wide format. * At least one release entry for What's New and Version History. Notis Apps are icon-led, so there is no separate cover image. App Details previews the same icon, screenshots, release history, and included resources that installers will see. Store details come from the deployed app. If the preview is stale, update the source, deploy that exact version again, and return to App Details. ## Write release history Keep the full history in the app's `CHANGELOG.md` file, with the newest release first. The first entry becomes **What's New**. The complete file becomes **Version History**. Use this shape: ```md theme={null} ## [A Better Daily View] - {PR_MERGE_DATE} - Added a faster way to review today's work. ## [Initial Release] - 2026-07-01 - Published the first version. ``` `{PR_MERGE_DATE}` is useful while a public release is waiting for review. Notis replaces it with the publication date when that release goes live. ## Choose who can install it On **App Details**, choose a visibility: * **Personal** — only you. Store publishing stays unavailable. * **Team** — members of your Notis team. * **Public** — anyone with a Notis account after review. Once a listing is live or waiting for review, visibility is locked. Unpublish or withdraw it before switching between Team and Public. ## Publish to your team 1. Choose **Team** visibility. 2. Make sure every Store checklist item is ready. 3. Select **Publish** on App Details. The listing appears in your team's Store immediately. After a later deployment, select **Update** to publish the next version. ## Publish to the public Store 1. Choose **Public** visibility. 2. Make sure every Store checklist item is ready. 3. Select **Publish** on App Details. 4. Wait for the Notis review. App Details shows the submission state: * **Pending review** — waiting for the Notis team. * **Approved & merged** — live in the public Store. * **Changes requested** — update the app, deploy again, and resubmit. * **Rejected** — review the note before deciding whether to resubmit. If you spot an issue while a submission is pending, choose **Unpublish** to withdraw it. You can submit a corrected version later. ## Publish with your coding agent The App Details button is the clearest review point. A coding agent can also submit the already deployed version after you explicitly confirm that the current Store preview is ready: ```bash theme={null} npx --package @notis_ai/cli@latest -- notis apps publish --confirm-ready ``` Deploying alone never publishes to the Store. The separate confirmation keeps an app from becoming visible outside your workspace by accident. ## Update or unpublish After another deployment, select **Update** on App Details. Team updates go live immediately; public updates go through review. From the **⋯** menu: * A pending public submission is withdrawn immediately. * A live Team listing disappears from the Team Store immediately. * A live Public listing stays available until its removal review is approved. People who already installed the app keep their copy after unpublishing. Unpublishing only stops new Store installs. ## Related * [Overview](/apps/overview) * [With Notis](/apps/build-with-notis) * [With your coding agent](/apps/build-with-your-agent) * [Install](/apps/install-from-store) # Changelog Source: https://help.notis.ai/changelog Product updates and announcements from Notis. ## July: Your Computer, and Notis for Coding Agents 07.26 cover Flo [here](https://x.com/mindtheflo) — last month was about control: intelligence levels, on-demand usage, social scheduling. This month is about reach. Until now, Notis lived in the cloud — its own computer, its own skills. Today it breaks out. Two features, one idea: meet you where the work actually happens. The first brings Notis **onto your own machine** — your shell, your local tools, the desktop apps that never had an API. The second goes the other way: it puts Notis **inside the coding agents you already use** — Codex, Claude Code, Cursor — so the skills and integrations you set up once follow you everywhere. Let's get into it. 07.26 image 1 — Your Computer ### Your Computer The Cloud Computer we shipped in v3 gave Notis a machine of its own. But some things only live on *your* machine — a script in a folder, a CLI you've configured, an app with no API, a tool running on `localhost`. So we gave Notis a way in. With the desktop app, Notis can now act **on your computer**, three ways: * **Local Shell** — run shell commands on your machine: open files, run scripts, drive the CLIs you already have set up. * **Local MCP** — connect Notis to MCP servers running locally, either as a local command or on `localhost`. Your private, local tools become Notis tools. * **Desktop Use** — hand Notis your mouse and keyboard to drive desktop apps that have no API. If you can click it, Notis can. This is opt-in and it stays under your control. Notis always prefers the safe path first — a connected integration, a Composio tool, or the Cloud Computer — and only reaches for your machine when that's what the job needs. You set the approval level, from **ask me every time** all the way to **run everything**, and you can dial it back whenever. [*Learn more about Your Computer.*](https://help.notis.ai/your-computer/overview) 07.26 image 2 — Coding Agents ### Notis for Coding Agents If you live in Codex, Claude Code, or Cursor, you know the tax: every new agent means re-teaching it your skills and re-connecting every integration from scratch. We think you should set that up **once**. Connect Notis once, and your coding agents work through your Notis account: * **Sync your skills** — author a skill a single time and keep it available in Codex, Claude Code, Cursor, *and* Notis. Update it in one place and the change follows everywhere. * **Share your integrations** — Gmail, Slack, GitHub, Notion, Linear, Calendar: connect them once through Notis instead of wiring each agent up separately. Your agents get the access; you skip the setup. * **Search your memory** — your agents can call `search_memories` straight from the CLI, so Codex or Claude Code can pull from everything Notis remembers about you and your work — no copy-pasting context, no re-explaining yourself. It all runs through the **Notis CLI**: install the desktop app, authenticate, and your built-in skills sync straight into your local agents. [*Learn more about Notis for coding agents.*](https://help.notis.ai/agents/overview) 07.26 image 3 — Now on GPT-5.6 ### Now running on GPT-5.6 Two features above — one upgrade underneath. In June we gave you the intelligence dial: Low, Medium, High, and Auto. This month we upgraded what sits behind it. Notis now runs on **OpenAI's brand-new GPT-5.6**: * **Low** runs on **GPT-5.6 Luna** — quick and light for everyday work. * **Medium** and **High** run on **GPT-5.6 Sol**, OpenAI's newest frontier model — dialing up the reasoning as the problem gets harder. Nothing changes in how you use it: the dial is the same, and **Auto** still picks the right level for you. The difference is that every task now runs on a sharper, more capable model underneath. ### 🐞 Other improvements * **Usage, per thread:** the Manager now shows how much each thread has used — so you can see exactly where your plan is going, right next to the work. * **Responsive follow-ups:** send Notis another message while it's mid-task and it now picks it up right away, instead of making you wait for the first one to finish. * **Smoother trials:** during a free trial, your plan details and the activate button now show up exactly where you'd expect them. * **Memories uploads, fixed:** dropping files onto your Memories page works reliably again. * **A steadier month:** plenty of behind-the-scenes fixes across voice, billing, and automations. ### 🗞️ In Other News * We shipped a brand-new [Integrations directory](https://notis.ai/integrations) this month — browse everything you can plug into Notis and see how to automate more of your work. * Two things that resonated this month: on LinkedIn, ["they have 2,000 paying teams and raised \$80M; we have 1,000 and raised \$0 — does efficiency beat funding?"](https://www.linkedin.com/feed/update/urn:li:activity:7476609773384400897/), and on Instagram, [a reel on ADHD, dopamine, and outsourcing the tasks you can't force yourself to do](https://www.instagram.com/reel/DaRGUdIDFiQ/). * I post daily across [X](https://x.com/mindtheflo), [LinkedIn](https://www.linkedin.com/in/parisetflorian/), and [Instagram](https://www.instagram.com/mindtheflo) — come say hi. ### 📰 **AI News** Here's what you might have missed in AI since June: 🔥 **OpenAI ships GPT-5.6 in three tiers — and it trained itself** — OpenAI released GPT-5.6 as Sol, Terra, and Luna (\$5 / \$2.50 / \$1 per million input tokens), with Sol hitting 91.9% on Terminal-Bench 2.1 and 90.4% on BrowseComp, plus a million-token context window. It also folded Codex and ChatGPT Work into a single desktop "superapp." Most striking: OpenAI says Sol autonomously post-trained the smaller Luna model with "limited human hands on the wheel." ([openai.com](https://openai.com/index/gpt-5-6/)) 🎙️ **GPT-Live makes AI voice full-duplex** — OpenAI launched GPT-Live and GPT-Live-1 mini, which listen and talk at the same time and quietly delegate hard reasoning to GPT-5.5 in the background. The new voice model scores 75.2% on BrowseComp versus just 0.7% for the old Advanced Voice Mode, and 84.2% on graduate-level GPQA. Live on iOS, Android, and web globally. ([openai.com](https://openai.com/index/introducing-gpt-live/)) 🎭 **Anthropic ships Sonnet 5 and brings Fable 5 back on a leash** — Claude Sonnet 5 landed at the same \$3/\$15 pricing (promo \$2/\$10 through Aug 31) but benchmarks below Opus 4.8, signaling restraint over raw capability. Meanwhile Fable 5 — pulled June 12 over export controls — returned July 1 wrapped in a new safety classifier that blocks cyber-exploit jailbreaks with 99%+ effectiveness. ([anthropic.com](https://www.anthropic.com/news/redeploying-fable-5)) ⚖️ **Apple drags OpenAI into federal court** — Apple filed a trade-secrets suit in the Northern District of California, alleging OpenAI poached 400+ former Apple employees — including 24-year veteran Tang Tan, now OpenAI's chief hardware officer — and built its devices on stolen iPhone IP. Apple wants a jury trial and a forced redesign of OpenAI's hardware. ([bloomberg.com](https://www.bloomberg.com/news/articles/2026-07-10/apple-sues-openai-for-trade-secret-theft-in-blockbuster-case)) 🧠 **Meta can now read your mind — no implant required** — Meta's Brain2Qwerty v2 decodes typed sentences straight from brain activity using an MEG scanner instead of surgical electrodes, hitting 61% word accuracy (78% for the best participant) — a leap from the \~8% earlier non-invasive methods managed. ([ai.meta.com](https://ai.meta.com/blog/brain2qwerty-brain-ai-human-communication)) 🛡️ **The EU quietly greenlit "Chat Control 1.0"** — The European Parliament let voluntary suspicionless scanning of private messages stand until 2028: the motion to reject fell short at 314 votes (361 needed). It lets platforms like Instagram, Discord, Snapchat, Gmail, and iCloud scan the private messages of roughly half a billion people without warrant or suspicion. ([patrick-breyer.de](https://www.patrick-breyer.de/en/eu-parliament-greenlights-chat-control-1-0-breyer-our-children-lose-out/)) 🏛️ **South Korea bets \$880 billion on the AI stack** — President Lee Jae-myung, alongside Samsung and SK Hynix, committed at least \$880B to semiconductors, data centers, physical AI, and robotics, framing it as national "survival." SK Hynix had already crossed a \$1 trillion valuation in May. ([bbc.com](https://www.bbc.com/news/articles/c9q2pwzngjqo)) 🦾 **Europe's biggest robotics round yet** — Germany's NEURA Robotics raised up to \$1.4B in the largest Series C ever for full-stack robotics, at a \~\$7B valuation, backed by Nvidia, Amazon, Qualcomm, and Bosch. With a \$1B+ order book for its 4NE-1 humanoid, NEURA is aiming to ship millions of robots by 2030. ([neura-robotics.com](https://neura-robotics.com/record-series-c/)) That's it for July. v3 gave Notis a computer and skills of its own — this month it steps onto yours, and into every agent you already run. See you next month. Flo P.S.: Record [a video testimonial](https://love.notis.ai/) (in any language) and get a free month of Notis on us. ## June: Intelligence Controls, On-Demand Usage, and Social Scheduling 06.26 cover Flo [here](https://x.com/mindtheflo) — after the v3 paradigm shift in May, June was about giving you the controls. v3 made Notis dramatically more capable. This release makes that capability *yours to dial in*: choose exactly how much intelligence each task gets, never hit a wall when you need more, and let Notis run your social calendar end-to-end. Underneath, it was also a heavy month of reliability work to make the new v3 surfaces rock-solid. Let's get into it. 06.26 image 1 ### Intelligence Controls Most AI agents drown you in knobs — a dozen models, a pile of "effort" settings, endless tradeoffs you were never meant to make. We think that's backwards. So we ran the evals for you and boiled it down to **two dials: intelligence and speed**. Every other optimization, we made for you. There are three levels: * **Low** — fast, light, and **up to 4x cheaper**: routine work now runs at roughly a quarter of the cost, so the same plan stretches up to 4x further before you hit your cap. (Hitting that wall was the single biggest reason people churned — so this one really matters.) * **Medium** — the balanced default for everyday work. * **High** — our most capable reasoning, for the problems worth thinking hard about. And on **Auto** — the default — there's nothing to choose at all. Notis reads each task and picks the right level for you, every time. Prefer to drive? **Pro+** users can override intelligence and speed **per thread** in the apps, **per channel**, and **per automation**, right from the Manager. [*Learn more about intelligence levels & rates.*](/account-and-settings/billing/usage-and-rates) 06.26 image 2 ### On-Demand Usage Running out of credits mid-task is the worst. So we killed the wall. With **on-demand usage**, you can keep going past your plan's limit and simply post-pay for the extra — no upgrade dance, no blocked automation at the exact moment you needed it to fire. You stay in control with usage warnings and clear thresholds, and a new **billing overview** in your settings shows exactly where you stand. For teams, credits are now **pooled across the whole team** instead of being trapped per-seat — so the person who needs them most can actually use them. [*Learn more about on-demand usage.*](/account-and-settings/billing/on-demand-usage) 06.26 image 3 ### Social Scheduling Notis can now run your social presence end-to-end — and it's **included**. Most social schedulers — Buffer and the others you'll find in the [Social Media guide](/guides/social-media) — charge **\$20–80/month** just to queue posts. With Notis it's **free, inside the license you already have**. We integrated **PostForMe** natively, so connecting X, LinkedIn, Instagram, TikTok, YouTube and more is a **one-click setup** in the Integrations tab — **no extra tools, no separate subscription**. Then just talk: dictate an idea on your walk, let Notis write and adapt the copy per platform, attach your media, and **schedule a whole week** of posts from a single message. [*Learn more about social media via Notis.*](/guides/social-media-via-notis) ### 🐞 Other improvements * **Safer sign-in:** Strangers can't trigger Notis or quietly create accounts anymore, and signing in by email now uses a simple, secure one-time link. * **Granola memories:** Connect Granola and Notis automatically remembers what came up in your meetings (ULTRA). * **Refresh memories on demand:** Re-sync your connected memory sources whenever you want, right from the Memories page. * **Smarter, leaner tool use:** Notis picks the right tool for the job more reliably — and uses far fewer tokens when you have a lot of integrations connected. * **No more guessing:** When an extension isn't connected, Notis now tells you — instead of hallucinating its way through the task you handed off. * **Transcribe anything:** Send an audio file and Notis turns it into clean, usable text. * **Better web reading:** We upgraded how Notis reads and pulls content from the web, so research comes back cleaner and more complete. * **Inline media in the Manager:** Images and files now show up right inside your Manager threads. * **In-app support:** A new **Contact Support** option opens a chat with us without leaving Notis. * **A steadier month:** Lots of behind-the-scenes fixes across voice, WhatsApp, billing, and automations so everything from v3 just works. ### 📰 **AI News** Here's what you might have missed in AI since v3: 🔥 **Anthropic drops Claude Fable 5 (and its caged twin, Mythos 5)** — On June 9 Anthropic shipped Fable 5 publicly while keeping the lightly-safeguarded Mythos 5 to trusted partners. Fable 5 took #1 on FrontierCode Diamond at 29.3% (vs Opus 4.8's 13.4% and GPT-5.5's 5.7%), priced at \$10/M input and \$50/M output. The system card flags Mythos at a "CB-1" biosafety rating creeping toward CB-2 — the kind of warning label that makes you read the rest of the page. ([anthropic.com](https://www.anthropic.com/news/claude-fable-5-mythos-5)) 🤖 **Claude is now building Claude** — Anthropic says 80%+ of production code merged into its codebase is now Claude-authored, with engineers merging 8x more code daily than in 2024. The kicker: Mythos Preview delivered a 52x model-training optimization speedup (vs 3x for Opus 4 a year earlier) and out-predicted human researchers' next move 64% of the time. Recursive self-improvement has left the whiteboard. ([anthropic.com](https://www.anthropic.com/institute/recursive-self-improvement)) 🏛️ **Dario Amodei demands binding AI rules** — One day after the Fable 5 launch, Anthropic's CEO published a five-part "Policy on the AI Exponential" calling for mandatory FAA-style third-party testing of frontier models, plus government power to block unsafe ones — and pledged "substantial financial backing" to push legislation. A notable swerve from the industry's usual transparency-only crouch. ([darioamodei.com](https://darioamodei.com/post/policy-on-the-ai-exponential)) 💸 **OpenAI eyes an IPO, ships GPT-5.5, and starts a price war** — Altman told staff OpenAI expects to go public "within the next year" (SEC paperwork filed, staff tender at \$687.69/share), with model 5.6 teed up for June. Meanwhile GPT-5.5 Instant became the ChatGPT default for hundreds of millions, cutting hallucinations 52.5% on high-stakes prompts and jumping AIME 2025 from 65.4 to 81.2 — as both labs reportedly weigh "drastic" token price cuts ahead of their IPOs. ([theinformation.com](https://www.theinformation.com/briefings/exclusive-openai-preps-new-ai-model-expects-go-public-within-next-year)) 🛡️ **Frontier models are getting scarily good at cyber** — The UK's AI Security Institute says frontier cyber capability has doubled every 4.7 months since late 2024, with Mythos Preview and GPT-5.5 now blowing past the trend (Mythos cracked 6/10 runs of a simulated network attack). On the upside, Anthropic's Mythos surfaced 271 Firefox security flaws for Mozilla in two months — same superpower, friendlier hat. ([aisi.gov.uk](https://www.aisi.gov.uk/blog/how-fast-is-autonomous-ai-cyber-capability-advancing)) 🦾 **BMW puts humanoids on the line** — BMW will deploy Hexagon Robotics' Aeon humanoids (1.65m, 60kg, 21 sensors, 15kg lift) in European manufacturing starting summer 2026 after Leipzig trials, feeding parts and doing battery pick-and-place. Trained via teleop, NVIDIA digital twins, and RL — the "robots take the factory floor" era is now a press release, not a prediction. ([bbc.com](https://www.bbc.com/news/articles/cgmpwzzvxr2o)) 🎯 **Reality-check benchmarks say agents still aren't there** — Two harder benchmarks landed: Workspace-Bench ran 28 agent setups on 388 real office tasks (best: Opus 4.7 + OpenClaw at 68.7% vs an 80.7% human baseline; average just 47.4%), and DeepSWE's 113 tasks average 668 lines of code per fix. Translation: coding and office agents are improving fast but still trail humans on messy real work — which is exactly the gap tools like Notis live in. ([deepswe.datacurve.ai](https://deepswe.datacurve.ai/blog)) That's it for June. v3 gave Notis its new shape; this month makes it yours to tune. See you next month. Flo P.S.: Record [a video testimonial](https://love.notis.ai/) (in any language) and get a free month of Notis on us. ## v3: Desktop App, Skills, Cloud Computer, and Memory Connectors 05.26 cover Flo [here](https://x.com/mindtheflo) — v3 is here, and this one is different. In February I teased "a paradigm shift for Notis". Then went quiet for 10 weeks. This is what I was building. Every previous release added a capability. v3 changes the shape of Notis itself. Three paradigm shifts in one: a desktop experience, a cloud computer, and new memory connectors. Don't be fooled: behind the scenes, this release sets the stage for the future of Notis — one not built on top of Notion. Let's get into it. 05.26 image 1 ### Notis Manager — Web & Desktop App The genius of Notis, if I dare say, has been to meet you where you already work most: on your messaging apps and your inbox. But as Notis became more capable, I found myself struggling to keep up just being the human in the loop - especially when it came to automations. Today we introduce the Notis Manager, a keyboard-first app designed to help you stay organized and validate your agent work systematically when you're back at your desk. Think Superhuman meets Claude: the inbox discipline of the best email client in the world, built to manage a swarm of agents for modern operators. **Inbox — Human-in-the-Loop** The real point of being the human in the loop is to validate—vet—the work done by your AI agents. For that, you need visibility into what's happening. The Manager provides this: when an agent completes a task, when a reply is needed, or when an automation reaches a decision point, you'll see it clearly in your **Inbox**. Review what matters, make the call, and archive. For messaging channels, Notis intelligently separates conversations by topic so you can easily transition from your favorite messaging apps to our new desktop experience. Automations are displayed together so you finally have full visibility of what Notis is doing behind the scenes. The framework is now complete: **Automations** are *when* — the triggers. **Skills** are *how* — the business logic. **Channels & integrations** are *where* — the destination. This tab ties it all together. Notis Manager inbox — weekly social scheduling automation marked done *The Inbox after a completed automation: Notis ran the weekly social scheduling skill, surfaced the posts for review, and is waiting for your call.* **Notis Overlay & Voice Mode** Hit **option + space** to summon Notis anywhere on your desktop. Hit **cmd + tab** to switch to voice mode and talk live to Notis with access to all your context, memory, and extensions. Voice mode used to be reserved for ULTRA — it's so good I've decided to bring it to **PRO+** starting today. Notis Overlay on desktop — option + space to summon *Hit option + space from anywhere on your Mac. Your full context, memory, and extensions — one keystroke away.* 05.26 image 2 ### Support for skills and your Computer in the Cloud This one is big. Skills has quickly become the standard for how to 'teach' agents your business logic. If automations are the "when", skills are the "how". Today we're launching full support for skills inside Notis. A skill is a markdown file and some scripts bundled together that Notis can invoke on demand. We ship a growing library out of the box: * **Word, Excel, PowerPoint:** Generate, edit, and export full Office documents — not just text — right from a message. * **PDF:** Create and edit rich PDFs with real layout and formatting. * **Browser Agent:** Navigate the web, fill forms, extract data, take screenshots — fully automated. Many upcoming capabilities of Notis will ship as out-of-the-box skills — but more importantly, you can ask Notis to create your own skills and edit them right in the portal. We also integrated skills.sh, the biggest library of skills, directly in the search bar. Here are some of my own: * **blog-post:** drafts a full blog post from a one-line brief, matches my tone, and publishes to Notion. * **social-media-weekly-review:** pulls my engagement stats, surfaces what worked, and suggests next week's angles. * **social-media-posting:** takes a draft and schedules it across X, LinkedIn, and Instagram in one shot. **Cloud Computer** To support skills, we equipped Notis with its own cloud computer — unlocking a range of capabilities previously reserved for locally run agents like Claude Code or OpenClaw. Notis can now open files, run scripts, install CLIs, launch browser agents, manipulate spreadsheets, render presentations, and more. 05.26 image 3 ### Memory Connectors (ULTRA) Notis already remembers what you tell her. Now she can remember your *entire world*. For **ULTRA** users, you can connect external sources — Notion, Gmail, Google Drive, OneDrive, Granola, websites — as **memory sources**. Notis indexes them continuously in the background and uses them automatically when relevant, without you having to mention them. **Memory sources vs. integrations — what's the difference?** * **Integrations** are active: Notis can use tools exposed by those integrations — send an email, update a database, book a calendar slot. * **Memory sources** are passive: Notis indexes them so she *knows* what's in there, even when you don't think to bring it up. Example: you ask Notis to write a new blog post. Without memory sources, she searches what you've explicitly written before to understand your tone. With memory sources, she already knows your tone, how you differentiate your product, and the doc you wrote six months ago — and synthesizes all of it automatically. ### 🐞 Other improvements * **⚠️ Multiple channel accounts:** You could already connect multiple accounts per integration — Gmail, Slack, Calendar. Now you can connect multiple accounts of the *same* service — two Gmail addresses, two Slack workspaces, personal and work calendar — all active simultaneously. Notis routes each correctly. * **Magic link sign-in:** Replaced our previous authentication with a more secure one-time link flow. Sign-in just works, without the brittleness. * **Bigger context window for threaded channels:** Slack, desktop and email conversations now get a larger context window by default — because threads deserve full context, not truncated history. * **Nano Banana Pro:** Image generation model upgrade — better quality, more consistent output, faster renders. * **Tool catalog in the Manager:** Browse every tool available to Notis — organized by category and source — directly from the Manager search bar. * **Dev mode:** Inspect raw tool calls and outputs inside the Manager. Useful for debugging automations and skills. ### 🗞️ In Other News * Back in February I wrote *"I'm also working behind the scenes on a paradigm shift for Notis."* That was 10 weeks ago. Now you know. 🤐 * Trial credits expanded from **\$5 to \$20** — new users have much more room to explore Notis before committing to a plan. * Comparison pages are going live: Notis vs OpenClaw, Notis vs Memorae, Notis vs Poke. * We're at [**\$16k MRR**](https://trustmrr.com/startup/notis-ai) — every one of you made that happen. Thank you. 🙏 * I post daily across [X](https://x.com/mindtheflo), [LinkedIn](https://www.linkedin.com/in/parisetflorian/), [Instagram](https://www.instagram.com/mindtheflo) — come say hi. That's it for v3. This is the Notis I set out to build — the one that doesn't just answer you, but actually does the work. See you on the other side. Flo P.S.: Record [a video testimonial](https://love.notis.ai/) (in any language) and get a free month of Notis on us. ## February: Notis Reliability Month, Interactive Buttons, and Patch Mode 02.26 cover Flo [here](https://x.com/mindtheflo) — this month, no big headline feature. I spent February making Notis more reliable, more predictable, and harder to break. If you delegate real work to an AI, "almost works" isn't good enough. That said, I’m also working behind the scene on a paradigm shift for Notis. 🤐 Let’s jump in. ### Interactive Buttons on Slack & Telegram Notis no longer just talks at you — now it gives you things to tap. When Notis presents options, links, or next steps, they show up as clean interactive buttons instead of raw text you have to parse and reply to. It's a small shift that makes Notis feel less like a chatbot and more like an app you actually want to use. Example: * Notis asks you to approve a draft — you tap "Approve" or "Edit" instead of typing it out. * Notis shares three calendar slots — you pick one with a tap. * A research summary lands with a "Save to Notion" button right below it. Reliability Month ### The Boring Release (a.k.a. Reliability Month) This is the real story of February. We shipped **30+ bug fixes** across the entire stack — concurrency, integrations, automations, messaging, and Notion sync. Here's why this matters: when you tell Notis to "post this at 9 AM" or "update my CRM after every call," you need it to actually work — every time. Not 95% of the time. Every time. What we fixed: * **Concurrency & locks:** Eliminated race conditions where parallel tasks could kill each other. Background threads no longer leak. Stalled processes get detected and cleaned up automatically. * **Integrations & OAuth:** Auth flows are faster and more resilient. Toolkit connections handle edge cases (multiple configs, first-attempt failures) without breaking. * **Messaging:** Telegram buttons no longer collide on long messages. Slack formatting is cleaner. iMessage media pipeline is more stable. * **Notion sync:** Database parsing, routing, and backfill logic are all tighter. Fewer edge-case failures during heavy usage. We also added **cost caps per agent** and **stalled interaction tracking** — so if something does go wrong, you know about it immediately instead of discovering it three days later. This is the kind of work that doesn't make a good demo, but makes everything else possible. Patch Mode for Notion ### Patch Mode for Notion Documents Sometimes you don't want a full rewrite — you want a surgical edit. With patch mode enabled, Notis can update only the relevant parts of a Notion page without bulldozing everything else. This matters most for long-living docs (plans, specs, playbooks) where stability is the whole point. You want iteration without "delete and recreate" chaos. Example: * "Add a Risks section to this doc" becomes an insert where it belongs, instead of rewriting the page. * "Update the intro with today's context, keep the rest" stays localized to the intro. ### 🤖 Automation of the Month: Competitors Comparison This month's featured automation tackles one of the most tedious tasks in content marketing — writing detailed competitor comparison blog posts. Instead of spending hours researching, formatting, and drafting, you just send Notis a competitor name and it handles the rest. **What it does:** This webhook automation takes a competitor name as input and runs a full pipeline: deep research on both the competitor and Notis (features, pricing, integrations, reviews, limitations), scrapes product illustrations from both websites, generates a custom header image, and upserts a complete, publish-ready blog post into your Notion content database — structured with a quick take, product overviews, a head-to-head comparison table, key differences, persona-based recommendations, and a CTA. **Perfect for:** * Founders and solopreneurs who need SEO-driven comparison content but don't have a content team * Marketing teams that want to scale competitor pages without manual research * Anyone who wants a structured, fair, and visually rich comparison post — ready to publish **→** [Watch the tutorial](https://youtu.be/PNjgpK7DvRI) ### 🐞 Other improvements * \*\*Better file manipulation: \*\*A thousand tiny improvements to facilitate how Notis manipulates files between the extensions, code interpreter and it’s context window. * **iMessage next-gen migration:** the iMessage channel now supports RCS, Whatsapp and SMS - all from the same phone number. * **Shorter onboarding:** Getting to value faster with less setup friction. * **Unified tool search:** Agents find and use the right tool more consistently. ### 🗞️ In Other News * Notis starts the year on the [TAAFT 2026 leaderboard at #8](https://www.linkedin.com/feed/update/urn:li:share:7432137363314864128) — overtaking CodeRabbit and Grok. Take this, OpenClaw! 😉 * I posted ["It's just a wrapper."](https://x.com/i/web/status/2022731729056801065) and it hit a nerve. If you know, you know. 🫡 * We now post daily across platforms — catch the latest on [X](https://www.linkedin.com/company/mindtheflo/), [Linkedin](https://www.linkedin.com/company/mindtheflo/), [Instagram](https://www.instagram.com/notis_ai), and [TikTok](https://www.tiktok.com/@notis_ai?_r=1&_t=ZN-94Itt5xvh7m). ### 📰 **AI News** Here's what you might have missed in AI this month: 🔥 **Matt Shumer: "Something Big Is Happening"** — The essay that broke the internet. 80M+ views in days. Shumer compared this AI moment to February 2020 before COVID hit, arguing we're in the "this seems overblown" phase of something much bigger. He claims he's "no longer needed for the actual technical work of my job." ([fortune.com](https://fortune.com/2026/02/11/something-big-is-happening-ai-february-2020-moment-matt-shumer/)) 💰 **Anthropic raises \$30B and Open AI \$110B** — Claude Code hit \$1B ARR within seven months. Meanwhile, OpenAI is targeting \~\$600B in compute spend by 2030, projecting \$280B in 2030 revenue. ChatGPT now has 900M+ weekly active users. The arms race is fully financialized. ([anthropic.com](https://www.anthropic.com/news) /[openai.com](https://openai.com/index/scaling-ai-for-everyone/)) 🇨🇳 **The AI Cold War heats up** — Anthropic accused DeepSeek, Moonshot AI, and MiniMax of "industrial-scale" distillation attacks: 24,000 fake accounts, 16M+ exchanges to extract Claude's reasoning and coding capabilities. OpenAI made parallel accusations to Congress. DeepSeek reportedly trained models on Nvidia Blackwell chips, flouting export controls. The framing from both labs: this is a national security issue, not just IP theft. ([techcrunch.com](https://techcrunch.com/2026/02/23/anthropic-accuses-chinese-ai-labs-of-mining-claude-as-us-debates-ai-chip-exports/)) 🤖 **OpenClaw / Clawdbot goes nuclear** — An Austrian dev's open-source AI agent — originally named "Clawdbot" before Anthropic sent a trademark cease-and-desist — became the fastest-growing GitHub project ever (150K+ stars). It runs locally, connects to WhatsApp/Slack/email, and actually *does things* autonomously (reminds you of something?!). Creator Peter Steinberger joined OpenAI on Valentine's Day. Security researchers found 30,000+ exposed instances. Running it costs \$300-750/mo in API tokens (still find Notis expensive ?!). It's your new weekend project, your security nightmare, AND the reason to buy that Mac Mini for local inference - or there is Notis. ([ibm.com](https://www.ibm.com/think/news/clawdbot-ai-agent-testing-limits-vertical-integration)) 🧠 **GPT-5.3 Codex: "The AI helped build itself"** — OpenAI's Feb 5 release included a line in the technical docs that sent shockwaves: "GPT-5.3-Codex is our first model that was instrumental in creating itself." The Codex team used early versions to debug training, manage deployment, and diagnose evals. Recursive self-improvement is no longer theoretical — it's a line item in release notes. ([openai.com](https://openai.com/index/introducing-gpt-5-3-codex/)) 🔒 **Claude finds 500+ hidden vulnerabilities** — Anthropic's Frontier Red Team used Opus 4.6 to scan open-source codebases and discovered over 500 security vulnerabilities that had gone undetected for decades despite expert review. They then launched Claude Code Security, which scans codebases for vulns and suggests patches — sending cybersecurity stocks tumbling in the process. ([claudelog.com](https://claudelog.com/claude-news/)) 🕵️ **Trump blacklists Anthropic — OpenAI swoops in hours later** — Anthropic refused to lift its two red lines for the Pentagon — no autonomous weapons, no mass surveillance. Trump ordered every federal agency to cease using Anthropic and Hegseth designated it a "Supply-Chain Risk to National Security." The \$200M DoD contract is dead. Hours later, Altman announced OpenAI had struck a deal for those same classified networks — claiming the Pentagon agreed to the *exact same safeguards*. ([axios.com](https://www.axios.com/2026/02/27/anthropic-pentagon-supply-chain-risk-claude)) 🎯 **Sonnet 4.6 launches Feb 17** — Approaching human-level performance on computer use tasks (72.5% on OSWorld). Same Sonnet pricing. 1M token context in beta. Plus: Anthropic acquired Vercept (perception/interaction AI) and the Bun runtime team. The product velocity is relentless. ([releasebot.io](https://releasebot.io/updates/anthropic/claude)) That's it for February. No fireworks — just a sturdier Notis. See you next month with big news! Flo P.S.: Record [a video testimonial](https://love.notis.ai/) (in any language) and get a free month of Notis on us. ## January: Notis Enhancements, Slack Integration, Teams Support, and MCPs