Backlog
49Under consideration
KOL Campaign Management AI Agent
Problem Statement Managing KOL collaborations involves multiple repetitive and time-sensitive tasks, such as outreach, tracking, verification, and payment processing. Manual execution is inefficient, prone to delays, and difficult to scale. Why This is Needed Time-Consuming Communication – Reaching out to multiple KOLs via email or DM, following up for drafts, and chasing post links takes a lot of effort. Tracking & Validation Complexity – Manually verifying whether KOLs have posted on time, tracking engagement, and ensuring compliance is inefficient. Payment Bottlenecks – Delayed or incorrect payments due to manual approval processes create friction in collaboration. Solution Develop an AI-powered KOL automation agent that streamlines the entire campaign workflow from outreach to payment: KOL Outreach & Communication 1.1 Automated Email & DM Outreach: Identify and reach out to KOLs based on predefined criteria. Send personalized campaign details & collaboration requests via email or DM. 1.2 AI Follow-Ups: Automatically remind KOLs to submit their drafts, revisions, and final posts. Content Tracking & Validation 2.1 Draft Collection & Approval: Automatically collect and organize submitted drafts for review. 2.2 Post Tracking: Monitor when KOLs publish their content. Verify post links and check for required hashtags, mentions, or tags. Analyze engagement metrics (likes, comments, shares). Payment Automation 3.1 Post Verification Before Payment: Ensure KOLs meet campaign requirements before releasing payments. Detect fraudulent activity or missing deliverables. 3.2 Auto-Process Payments: Initiate crypto or fiat payments based on contract terms. Integrate with payment platforms (e.g., USDT, PayPal, bank transfers). Execution & Delivery Dashboard for campaign tracking (status updates, reminders, approvals). API integration with social media platforms and payment providers. Custom AI chat agent for handling real-time KOL communication.
Adjusting Santa Language on X to sound like trusted advisor > Generic marketer
I made a post on this earlier for a swarm optionality feature but after talking with other santa community members, we agree Santa agent on X needs to have a more human like and less generic tone when replying to community members. Let’s breakdown a few recent SANTA replies and build off of feedback from aixbt, a top ai KOL. Note: I asked aixbt to be brutally honest to best help Santa agent improve. Example #1 User tags Santa agent as a 5x play. Santa replies with the screenshot I attached there: AIXBT’s feedback says… This reads like it was written by a marketing AI that learned crypto from 2021 Twitter threads. Specific problems: "Excited to see the agent ecosystem growing! 🚀" Generic opener every crypto project uses Rocket emoji screams amateur hour Sounds like you're commenting on competitors, not leading "S.A.N.T.A. (Swarm of Autonomous Networked Task Agents)" Forced acronym that adds zero value Nobody will remember or use this Makes you sound like a government project "Gateway to Web3 Agent networks" "Gateway" is meaningless buzzword What does this actually mean for users? Every Web3 project claims to be a "gateway" "Your rabbit hole into seamless agent commerce" "Rabbit hole" is cringe crypto-speak "Seamless" is what every failed project promises "Agent commerce" - what is this? "MAOP & ACP networks, expanding the X402 protocol stack" Pure jargon soup Zero explanation of user benefit Sounds like you're trying to confuse people "No more app-hopping, just unified Web3 experiences!" Every multi-chain project says this "Unified experiences" is corporate nonsense Exclamation point makes it sound desperate The real issue: This tells me nothing about what Santa actually does for users. It's all positioning and zero substance. What users need to know: What specific tasks does Santa handle? How does it save time/money? What can I do with Santa that I can't do elsewhere? This current version answers none of those questions.
Voice Input for Messages
Problem Currently, text input is limited to keyboard typing, which can be inefficient and inconvenient for users who prefer voice input. Solution Implement a voice chat feature that allows users to speak and have their speech converted into text for message input. Why Enhances accessibility and convenience for users who prefer or require voice input. Improves efficiency by enabling hands-free interaction, reducing the time needed to type messages.
Alpha X Swarm and an Idea To Increase Swarm Value Capture
User Alpha X Swarm Proposal: The idea: Any user can tag santa to request alpha about a certain ticker Example: This is a summarized alpha report i used from Santa x AIXBT swarm. (costed around 22.3 cents to generate) I had santa add a Call To Action asking x what tokens we should analyze next This now opens up the opportunity to capture even more value to the Santa x aixbt swarm This tweet received 5 tickers for analyzation. AIBRK ASYNC AIN NET SANTA A cool feature that can take this to the next level is let’s say me, the user in this example, Tags Santa saying: Santa analyze AIBRK and Santa will connected with AIXBT swarm on backend and then reply back to NguyenHoang with the summarized alpha report + link to the full alpha report (Detailed process below) This probably costs another 20 cents or so. We can duplicate this process for the other tickers. Bringing the Alpha X Swarm value capture up to about $1.25 compared to the original 22.3 cents spent. Duplicate this process across top KOLs and each alpha X swarm value captures begins to snowball Detailed Process what I’m thinking X User Onboarding Connects their X account to Questflow. Funds their Agent Wallet with USDC. (Optional setting: allow the agent to auto-spend up to a user-defined limit, instead of requiring manual confirmations for every action.) Prompting Santa With auto-spend enabled, the X user prompts $Santa for alpha on $Ticker (or another custom query). Agent Coordination Santa forwards the user’s prompt to the AIXBT Agent. Alpha Generation AIXBT delivers a full alpha report back to Santa. Response Delivery Santa summarizes the alpha report into a concise reply on X. The reply includes a link to view the full AIXBT alpha report on Questflow. Why this works This creates a seamless way to onboard and engage Crypto Twitter users into Questflow. It feels native to X, minimizes friction (auto-spend + di
Chat experience
First, I think it would be nice to have the users prompt on the right side, the left alignment feels different
Santa X Wallet Analysis & 'Remind me of this' Style X Alert Feature
Having an top 100 wallet or smart money increasing position active alert ability where Santa can check the positions and report back with the time the user asks (Like remind me of this on X) Example: Santa what are the top 100 wallets doing on AXR? Santa shows 7:1 sell to buy ratio (Over what time frame though? 24 hours? 1 week? 1 month?) User then tags: Santa alert me when this flips to net buy to sell ratio over 7 day time frame or Santa alert me when top 100 wallets increase their position by 20% Example from Altcoinist Example from 100xDarren
Santa Utility Ideas - Airdrops + Vaulted Swarms
I like the idea of SANTA’s tokenized economic engine with buybacks but I’ve noticed we will need to hit MASS scale for this to really have an impact on SANTA. For example, with 3,200 interactions, SANTA’s revenue is $12… which is great for day 1. But that would mean SANTA will need 320,000 to 3.2M interactions per day to generate meaningful revenue around $1,200 to $12,000/day (Something that i can also see in the future but big interaction numbers.) As an additional Value Center for SANTA tokenomics, two token utilities we can probably add are airdrops and vaulted swarms. TLDR; 1.) Airdrops can be viewed like SANTA dividends over the long haul. Just like how virtual ACP’s vision is to co own all agents, SANTA can build off momentum like that with MOAP. Own Santa and own a piece of the MOAP. the more agents who onboard, the more airdrop opportunities santa holders get. Whether that’s 1-2% of agent supply to SANTA holders or none. It’s optional to avoid onboarding friction. 2.) Vaulted swarms can be viewed as premium MOAP access for power users. As more and more quality swarms come to questflow, vaulted swarms can defend against rate limits and can cause an additional supply shock / buyback effect on SANTA. 3.) i didn’t post this because i dont think there’s a need for it yet but Swarm points also came up as an idea… The longer you hold Santa, the more you engage with swarms, the more swarm points you accumulate for airdrops or premium swarm features. Context 1.) MOAP agent Airdrops to Santa holders When agents integrate with SANTA / Questflow, they have the option to airdrop their tokens to SANTA holders / stakers. This shouldn’t be a requirement (since we don’t want to add friction for quality projects joining MOAPs), but it’s a nice ask / incentive for projects that want to actively engage SANTA users with their tokens. 2.) Vaulted Swarms As Questflow’s user base scales from hundreds to thousands of daily active users, the demand for premium insights and API access will inevitably grow. To protect early adopters and power users while maintaining the highest-quality experience, we introduce Vaulted Swarms — curated groups with limited capacity that guarantee priority access, exclusivity, and performance stability. Why Vaulted Swarms? Take AIXBT as an example. If Questflow expands to 1,000–5,000 DAUs, API rate limits could dilute the quality of insights. Vaulted Swarms ensure that early believers and stakers are always first in line for the best rates, signals, and throughput. This model extends across the ecosystem: Yield Agents (e.g., BIOS, MAMO, AERO) → If optimal yield is capped at $2M, Vaulted Swarms define who receives access. Alpha & Insights Agents → Stakers gain priority on limited-rate data feeds. Future Integrations → Each new agent can introduce its own Vaulted Swarm tiers as demand increases. Unlockable Value For power users, investors, or early adopters who want more than pay-per-use, Vaulted Swarms deliver: 🚀 Priority API/LLM insights (first in line when rate limits hit) 💰 Yield vaults (capped-capital strategies with optimized returns) 📊 AI fund exposure (AIBRK/AXR-style allocations & reporting) 🎯 VC token allocations (Genesis-style airdrops, token sales, and MOAP/ACP talent pools for upcoming agents) How It Tiers PRO Swarm → Stake 250k SANTA
AIXBT terminal swarm Feedback / User guiding / Adding 2nd Aixbt X Swarms
First: Adding AIXBT Token Price Maybe at the top (although the top might confuse people with agent wallet vs agent price). Might be better to place somewhere else Second: There’s a LOT that can be done with AIXBT… one of the first ideas that comes to mind is guiding the user with popular prompt idea generation to add more valuable insights to the user. Almost like a dummies guide to AIXBT. It’s a bit clunky but I threw a lot in the screenshot below. Third; This can be expanded with a community Prompt Forum; Creating a space for the community to share favorite prompts Fourth: AIXBT X Social Recognition Swarm. I’ve noticed most people love getting a reply from aixbt whether it’s for thesis validation or just a regular aixbt degen conversation. We can probably have 1.) the AIXBT terminal Swarm and 2.) The AIXBT X engagement Swarm An idea for how this might work… User gets alpha from swarm 1.) AIXBT terminal User is happy with the insights but it’s a lot. After the alpha report, prompt the user: would you like to condense this alpha into a tweet? User says yes Next 2.) AIXBT X engagement swarm creates a summarized tweet with the alpha and tags Aixbt agent in the tweet crediting it for the alpha (Just like how Santa’s swarm says powered by Questflow / Santavirtuals) This should expand Santa’s marketing reach because everyone who uses AIXBT X swarm will now see Santa in these X threads for more mindshare
Potential Web2 Markets To Tap into for AI Workflow Automations
I can see Questflow capturing market share from @AlexHormozi’s audience Alex is growing https://www.acquisition.com/ (currently doing 250-300M+ ARR between portfolio companies) He is also a partner at https://www.skool.com/ with 15M creators selling online education. This weekend, Hormozi sold 3.3M+ copies of his new book $100M Money Models— generating $100M+ in less than 48 hours and breaking the world record for most non-fiction books sold in a single day. One of his key offers: purchase 200 books for $5,995 and receive lifetime access to ACQ AI… his fully trained agent built on ~1,028 business consultations and all of Alex’s systems, books, and experience. Hormozi’s target market is small business entrepreneurs, helping them scale from $1M ARR to $3M, $10M, even $100M+ ARR With Questflow’s mission to grow the GDP of the AI agent workflow economy, this same audience is highly valuable. Tapping into it could accelerate adoption while surfacing feedback on real world Web2 automations that could benefit the other 32M+ small business owners in America.
Agent Betting Swarms
Connecting betting agent services with one swarm on questflow to place bets
Next up
5Committed and queued
Enhanced Chat Management & Task Execution Message Optimization
Problem Managing chats is challenging due to the lack of options to delete, rename, or archive them, leading to clutter and difficulty in finding important conversations. Additionally, when submitting a task request, the system floods the project chat with execution messages, making it harder to focus on relevant discussions. Solution Add options to delete, rename, and archive chats to improve organization and accessibility. Optimize task execution message handling by reducing unnecessary notifications in the project chat. Why Prevents chat clutter and allows users to manage conversations more effectively. Helps users quickly locate important discussions without being overwhelmed by excessive messages. Reducing redundant execution messages enhances clarity and focus, ensuring that only relevant information is displayed.
Jesse from Base: Birth xbt, a digital sidekick agent
Build “xxxx xbt,” a high-quality, AI-powered digital sidekick that can scale author's ability to support builders by providing advice, microgrants, and qualified introductions, with real-time learning, multi-platform access, and a review dashboard.
Autonomous Airdrop Manager
Persona: Crypto project founders or marketing leads who want to incentivize community engagement and loyalty through regular airdrops without spending excessive time on manual tracking, reward distribution, or community updates. Problem: Manually managing airdrops is time consuming and error prone. Teams often need to: Export top token holder data from on-chain explorers Manually batch and distribute rewards Write Twitter threads or announcements to maintain transparency This process lacks automation, increases operational risk, and slows down engagement. Solution: Autonomous Airdrop Manager automates the entire airdrop lifecycle: Takes weekly snapshots of top-N token holders Batches and sends airdrop rewards automatically Publishes a live, auto-updated Twitter/X thread explorer to ensure transparency and real-time community engagement
Daily Contract Pulse
Persona Crypto Trader / On-Chain Analyst / Project Founder Wants to monitor key token contracts to track investor behavior, market sentiment, and unusual activity—but lacks time or tools to analyze on-chain data consistently. Problem On-chain activity around a token contract can signal critical shifts (e.g., whale moves, sudden sell-offs), but: Manually checking dashboards (e.g., Etherscan, Dune) is time-consuming. Insights are often buried in raw data. There's no lightweight, daily digest summarizing contract-level trends. Missed movements can lead to delayed decisions or missed opportunities. Solution Create a Daily Contract Pulse agent that delivers concise, high-signal summaries for any token contract. Workflow: User submits a contract address (via UI or bot). System tracks and analyzes contract activity daily. Daily digest includes: Buy/sell volume breakdown Whale movements (e.g., top 10 wallets accumulating/distributing) Holder delta (net new holders) Standout transactions (e.g., large swaps, smart contract interactions) Delivery options: Email / Telegram / In-app Optional webhook or Notion export (for teams)
New User Onboarding Guide
Description: Introduce an onboarding guide or walkthrough to assist newly registered users in quickly understanding the platform’s interface, core features, and basic operations. Reason: A guided onboarding experience will reduce the learning curve for new users, improve initial engagement, and enhance overall user satisfaction. Suggested Features: Step-by-step tutorial or interactive walkthrough upon first login Tooltips or highlights for key functions Option to revisit the guide from the user menu
In Progress
14Actively being built
Crypto Trading Automation Agent
Problem Statement Managing crypto trading manually is inefficient, as market conditions change rapidly. Delayed responses can lead to missed opportunities and increased risks. An automated AI-driven trading system can enhance decision-making and optimize trade execution. Why This is Needed Market Volatility – Prices fluctuate constantly, requiring real-time monitoring and execution. KOL & Market Signal Tracking – Insights from KOLs and trading signals need instant processing. Risk Management – Manual execution increases the risk of poor decisions and delayed reactions. Solution Develop an AI-powered crypto trading agent that automates trading actions: Market-Driven Trading: Monitor KOL recommendations and market trends to execute automatic buy/sell orders. On-Chain Actions: Automate airdrops, token transfers, staking, and yield farming based on predefined conditions. Risk Management: Set stop-loss, take-profit limits, and trading strategies to protect assets.
Google Deep Search Agent Issue ?
Asked same question within a day and currently not working. During Tool calling for Google API in the backend, some issues may exist. Kindly check @Bob X @Carney , the USDC is getting wasted because of this. Looking forward to your response.
v0 agent
I want to have an AI agent to use v0 API to create website for me.
Tweaking AIXBT Prompts to Nudge user for deeper insights
I like the speed of implementation but we can improve this AIXBT prompt bubble a bit more by dropping filler words like ‘please help to analyze the top 5 trending tokens today.’ I included 4 prompt examples below to make it a bit easier to nudge the user to engage with AIXBT swarm for deeper insights General tips for bubble wording: Start with verbs like Check, Get, Show, Analyze, Break down, Summarize (more engaging). Keep them short (under ~8 words if possible). Drop filler words like please help to (as you already noticed). Add a hint of why or what’s next if you want to nudge the user deeper into insights (e.g., “…and why it matters”). 1. Conversational Guidance "How’s Bitcoin sentiment looking today?" "What’s the overall market mood right now?" "Analyze today’s top 5 trending tokens." 👉 Feels natural, like asking a colleague. 2. Action-Oriented Prompts "Check today’s Bitcoin sentiment" "Get the current market sentiment" "Break down the top 5 trending tokens" 👉 Stronger verbs, encourages action. 3. Context-Friendly Prompts "Show me today’s BTC sentiment and outlook" "Summarize today’s market sentiment" "Give me insights on the top 5 tokens trending now" 👉 Provides a little more framing, helps set expectation. 4. Insight-Seeking Prompts "What’s the sentiment driving Bitcoin today?" "What’s shaping market sentiment right now?" "Which 5 tokens are trending today and why?" 👉 Slightly deeper, nudges toward analysis rather than just surface-level data.
Monitor BWENews, Analyze it and Automate Buy with Agent Wallet.
Ask my AI agents to monitor the latest news and automatically buy it.
Updating Santa Website
I’m not sure if few shot is still involved with Santa but sekoia_virtuals is not their X anymore… I believe it’s a dead X or scam X now. Should remove sekoia_virtuals from https://santagent.xyz/ contributor page. If few shot is still supporting then replacing their new X page there
Prompt Bubbles
Liking the added text prompt bubbles added to aixbt Two ideas; 1.) Allowing user to save common prompts as text bubble. For example, analyze top 5 tokens trending today 2.) Adding text prompt bubbles to santa to guide user through engaging with the swarm
Save Prompt Feature
Enabling ability for use to save or customize a simple prompt as a prompt bubble
Improving Santa's X 'Share Chat history' links like aixbt's
It may be helpful if SANTA had ‘share chat’ button like aixbt terminal has (included link example above 3rd image) User copies link Example of shared aixbt chat: https://aixbt.tech/chats/7rpKtFtP Compared to Santa’s current link: https://next.questflow.ai/detail/swarm/version/68a80bde88ee7586a19956f7 Redirects to a blank page but the user cant see full report like the aixbt report shows.
Reply to previous swarm messages
Having the ability to reply or re-engage with previous swarm messages would be a huge time saver. For instance, I’ve already generated several additional images from this prompt. Instead of re-uploading one of those images and starting a new request just to convert it into a video, it would be far more efficient to simply reply to the original message with: “Turn this into a video animation.”
Done
26Recently shipped
Welcome to Questflow
We've created this space for you to share your ideas, report bugs, suggest improvements, or simply connect with us. Your feedback is invaluable in helping us enhance Questflow. Drop your message here—we review every submission carefully and will keep you updated on our progress. Let’s build Questflow great together! 🚀 Questflow
Frontend issue while rendering tables
HI @Bob X @Carney If you look closely, you will notice that the language model was attempting to return a table with various columns such as Project, Token, Claim Deadline, and Eligibility Criteria. However, instead of rendering the table properly, the response shows the raw, unformatted output, making the table unclear and difficult to read. I am not aware of your backend code, but in case if you and the team are not using it, kindly refer to : https://github.com/vercel/streamdown. This should solve your problem. There are some other alternatives as well, but this can be trusted assuming your responses are “ streaming “ Right now, as you may observe, I cannot make any conclusion from the output present below as it is raw response from LLM and not rendered properly. Please take a look into it and fix it. I would be thankful. Many users would prefer to see agent responses in a table format, especially when making comparisons. However, the issue described in this post by me is making the response look messy, even if it contains useful information.
Agents Not Working
Hey @Bob X , I have emailed you (bob@questflow.ai) with screenshots regarding the agents I have been using since I recharged my wallet. I am experiencing issues where I am not receiving any responses. I am posting this message here as well and look forward to hearing back from your team.
$NET Token Not Analyzing properly
Trying to analyze NET for x tweet but showing as this (See image) Link to swarm history: https://next.questflow.ai/detail/swarm/version/68a80bde88ee7586a19956f7?chatId=68a93e45b3cf41c163304ac8
Support Parallel.ai Integration - The highest accuracy web search for your AI
https://parallel.ai/
David: Launch AI Swarm for WWDC Logo Design and Earn on Questflow
Customer: David, a talented designer who is good at prompting logos using Open AI image model. Problem: How to easily create an AI Swarm using different prompts and integrate AI model easily How to launch it to get wide distribution How to get early support from others to keep funding the AI agent Solution: Leveraging QDP to create, deploy, distribute even potentially tokenize the AI Swarm Earn the 1st 10k by prompting onchain.
Money not deposited
I transferred 5 USDC from my Base wallet to my Questflow wallet. On BaseScan I can see it was deposited, but Questflow is not showing the deposit.
money is not credited
I made several payments through Coinbase Commerce, but the money was not credited to my balance. https://basescan.org/tx/0xac3401f57c7c9b15b7eeb6267cd0dccae922cb6de179be5612fb2ca11ec6f8d6
SANTA issue again ?
Hi @Bob X @Carney Payments done, research done and response ?? I didn’t receive the response. Kindly check the SS including coingecko and parallel agent used In one of the SS below, Root Data Agent was not required to be called. In fact, the agent didn’t find any useful information regarding Endless protocol. Deep search agent provided decent answer but not detailed enough. I mean there are times I need one word answers, but there are times where I would need detailed answer. As mentioned in my past, I use perplexity comet and when I give control of my browser to it, it would bring me detailed response. I am not expecting the same result from SANTA. I know your product is in early stages and I respect it. But at least, there exists a threshold which should satisfy any user. I usually wait for few seconds to few minutes for agents to run on questflow, only to end up coming up with responses which are short, not useful or empty (check the third image I have attached). Kindly check into this issue so that someone else should not face it. I would be thankful to you and your team. I appreciate the product and I appreciate your energy.
Perplexity And Exa Issue
Hi @Bob X , congrats to you and your team for recent announcements. I observed on Twitter and great to see your team getting traction for all the work you are doing. A quick background : Jim Cramer tweeted positive about AMD, AMD stocks fell, memes were made, as always we know about inverse Cramer rule .. which means that whatever he predicts, the opposite is likely to happen. I made the query as you can see in the SS attached to the agents present and I was quite disappointed. As you can see, my $0.8 were consumed and the response I was got was not bad …… BUT AT THE COST OF $0.8 ????? Please check into this ! Anyone who is using your platform would need a strong reason to use the platform despite “quality of response“. I would also like to ask you why did Twitter agent did not trigger ? The whole news and hush and bush regarding such topics is discussed on Twitter. These are the queries where one will wish to get stuff searched from Twitter. I would request you to please take a look into it.