I review AI tools by building with them. Screen-recorded tests, scored features, straight verdicts. Community builder. Paid collabs only, DMs open.
AI
Joined September 2026
- Tweets664
- Following9
- Followers398
- Likes659
Pinned Tweet
I review AI tools by building with them.
No press-release rewrites. I sign up like a normal user, build something real, screen-record every step and track what it costs.
What you get in every review:
1. Setup time from sign-up to first result
2. Real builds, not demo clicks
3. Cost receipts, credit by credit
4. What works and what breaks, with screenshots
5. A straight verdict
Example: I tested Zite with 4 business apps and 4 prompts, on 8.8 of 50 free credits. The dashboard maths was exactly right, and the first click on the live order form did nothing. Both made the review.
Brands get my findings before I post, so they can fix issues first.
Tools I cover: AI agents, agent tools and APIs, automation, AI video and image generation.
Building an AI tool? Paid collabs only, DMs open.
CLEMENT AI retweeted
Yesterday we said: start every clip from one image of your character.
Here is how to make that image.
Facing the camera. Plain background. Even light. Calm face. Head and shoulders.
Who is your character? Describe them in one line and we will tell you what to add so the face stays the same.
Prompt in the first reply. Save this.
Your character has a new face in every clip?
Stop starting each clip from text alone. Make one clear image of them, then start every clip from that same image.
Same image, same description, every time.
Try it and reply with your result.
What changes most in your clips: the face, the hair or the outfit?
CLEMENT AI retweeted
Wherever you are in the world, you can learn AI with NEXTGEN.
Nigeria. Africa. America. Europe. Asia. Anywhere.
Learn how to:
• Create AI videos, films and UGC ads
• Build websites and apps with AI
• Automate YouTube, Instagram, TikTok, Facebook and other platforms
• Build autonomous AI workflows
• Create AI agents, Discord bots and Telegram bots
• Design graphics and build brands
• Turn AI skills into real opportunities
And you can learn it from scratch for FREE.
NEXTGEN is building a global community of creators and builders who learn, build, share and grow together.
You don't need to know everything before you start.
Join NEXTGEN. Stay active. Stay committed. Build with us.
CLEMENT AI retweeted
I gave an AI agent 2,000 credits and 90 real API call on @Nevermined_ai. It spent 708 credits, about 71 cents.
Disclosure: Nevermined provided the credits for testing. 2,000 credits is about $2 at their own pack price ($10 for 10,000). No card, no wallet, no top-ups. Everything below ran the way an agent would call it, through their REST API and MCP server.
What I tested
Easy: IP lookups, a weather check, geocoding.
Typical: a research agent across three search engines plus a scraper, a due diligence check on a real website, an SEC filing search, a DeepL translation, AI images, and auto-pick, including asks it can't do.
Hard: bad requests, wrong paths, oversized requests, the same request sent twice, an LLM call, a fake API key, a full MCP run, a one-prompt "US Creator Weekend Brief", a 40-call reliability loop, and a full ledger audit.
What it does well
➢ Billing is exact. Every charge in the ledger added up to my balance drop, to the credit. Every quote matched the charge.
➢ Reliability was near perfect. 88 of 89 valid paid calls came back. The one failure was refunded automatically and the retry worked. The reliability loop went 40 of 40.
➢ Failed calls are mostly fair. A bad request was refunded. Wrong paths, missing fields and oversized requests were refused before any charge.
➢ Duplicate protection works. Same request twice, one charge.
➢ The one-prompt test is the real win. Two search engines, a scraper, an LLM, DeepL, an image and file hosting: 8 calls, 93 credits, 79 seconds, zero vendor signups. It found 3 real upcoming US creator events, with a Spanish hook and a hero image.
➢ It's built for agents. The MCP server tells the agent its balance, the cost and when credits expire, in plain sentences.
➢ It says no honestly. "Book a hotel room" came back as no match, charged nothing.
What it does badly
➢ Slow. Every paid call takes about 4 seconds. The same IP lookup straight from the vendor took 0.2 seconds.
➢ Headline prices are off. DeepL's card says half a cent. I paid 5 cents, 10x the headline.
➢ The markup is hidden. On average I paid 1.68x what vendors received. The LLM call was about 24x.
➢ You pay for junk if the vendor says 200. A paywalled scrape, two useless SEC searches and a geocode that put a well known neighborhood in the wrong country were all charged. The brief's LLM also got one event date wrong.
➢ "No stablecoins" covers part of the shelf. Credits pay 135 of 160 services and 61% of endpoints.
➢ No spending cap on credits. A looping agent could drain the whole balance.
➢ Dashboard rough edges. It says "No purchases yet" while Purchases lists every call, and "Total spent" shows a blank dash.
What 1,000 credits (about $1) buys at measured prices
➢ 500 IP lookups
➢ 333 page scrapes
➢ 200 AI images
➢ 125 short GPT-4o-mini answers
➢ 100 Exa searches
➢ 20 DeepL translations
➢ 12 website screenshots
➢ About 10 full US creator briefs
Scores (my judgment)
Onboarding 8
Discovery 8
Agent experience 8
Billing accuracy 10
Reliability 9
Failed-call fairness 7
Price transparency 5
Latency 4
Coverage on free credits 6
Budget controls 4
Docs accuracy 6
Verdict: 7/10
The plumbing is the best I've tested. Exact money, almost nothing broke, and one key replaced 13 vendor signups. It loses points on speed, headline prices that don't match what you pay, and charging for useless "successful" answers.
➢ Use it if you're building an agent that needs many data sources without signing up for each one.
➢ Skip it if you need speed, heavy use of one vendor, or a hard spending cap without a crypto wallet.
What I'd fix first
Show the real per-call price on every card
Add spending caps for credits, not just USDC
Cut the 4 second overhead per call
Pass vendor error messages back to the agent
Fix catalog paths the router refuses
Let agents flag junk 200 answers for a refund
Would you let your agent spend money on its own? Tell me below 👇
CLEMENT AI retweeted
Your character has a new face in every clip?
Stop starting each clip from text alone. Make one clear image of them, then start every clip from that same image.
Same image, same description, every time.
Try it and reply with your result.
What changes most in your clips: the face, the hair or the outfit?
CLEMENT AI retweeted
A social media manager in Kigali runs pages for three small businesses. A coffee shop, a fashion boutique and a gym.
Every Monday he stares at three empty calendars. Each client needs fresh ideas, a different tone, and captions in English and French.
By Wednesday he is posting whatever he can think of that morning.
Here is how he could set up an AI agent to do the first draft of his week, in four steps.
Step one: write a short brief for each client. What they sell, who buys, their tone, and any offers or events this month.
Step two: open @CreaoAI and type, in plain words:
"Build me a content planner for my clients. I manage social media for a coffee shop, a fashion boutique and a gym. Here is a brief for each one: [paste briefs]. When I name a client, give me a 5-day plan with a post idea, the best format (photo, reel or carousel), a caption in English and a French version, and one question that gets people to reply. Keep each caption under 120 words. Never invent prices, offers or events I have not given you."
That message builds the agent. No coding.
Step three: each Monday, name a client and get the week's plan.
Step four: edit what does not sound right, take the photos, and schedule the posts.
He still knows his clients best. The agent just means he never starts the week from a blank page.
Explore CREAO with link in bio to see how this works you can build any agent of your choice if you can describe the task CREAO can do it for you.
Creators and social media managers, how many pages do you run? Reply with your number and your country.
Save this for your next Monday. Run one page or ten, the agent works the same way. Reply with the kind of business you post for and I will write the brief for step one with you.
CLEMENT AI retweeted
Your phone photo of a product can look like a studio shot.
One prompt: plain background, soft light, sharp label. Swap in your product and run it.
Prompt is in the first reply. Save this.
What are you shooting first? Tell us the product and we will suggest a background to try.
CLEMENT AI retweeted
If you are building and you need an honest review, feedback, suggestions and recommendation.
DM Let's test what you're building or drop your product link below
I review AI tools by building with them.
No press-release rewrites. I sign up like a normal user, build something real, screen-record every step and track what it costs.
What you get in every review:
1. Setup time from sign-up to first result
2. Real builds, not demo clicks
3. Cost receipts, credit by credit
4. What works and what breaks, with screenshots
5. A straight verdict
Example: I tested Zite with 4 business apps and 4 prompts, on 8.8 of 50 free credits. The dashboard maths was exactly right, and the first click on the live order form did nothing. Both made the review.
Brands get my findings before I post, so they can fix issues first.
Tools I cover: AI agents, agent tools and APIs, automation, AI video and image generation.
Building an AI tool? Paid collabs only, DMs open.
If you are building and you need an honest review, feedback, suggestions and recommendation.
DM Let's test what you're building or drop your product link below
I review AI tools by building with them.
No press-release rewrites. I sign up like a normal user, build something real, screen-record every step and track what it costs.
What you get in every review:
1. Setup time from sign-up to first result
2. Real builds, not demo clicks
3. Cost receipts, credit by credit
4. What works and what breaks, with screenshots
5. A straight verdict
Example: I tested Zite with 4 business apps and 4 prompts, on 8.8 of 50 free credits. The dashboard maths was exactly right, and the first click on the live order form did nothing. Both made the review.
Brands get my findings before I post, so they can fix issues first.
Tools I cover: AI agents, agent tools and APIs, automation, AI video and image generation.
Building an AI tool? Paid collabs only, DMs open.
CLEMENT AI retweeted
I review AI tools by building with them.
No press-release rewrites. I sign up like a normal user, build something real, screen-record every step and track what it costs.
What you get in every review:
1. Setup time from sign-up to first result
2. Real builds, not demo clicks
3. Cost receipts, credit by credit
4. What works and what breaks, with screenshots
5. A straight verdict
Example: I tested Zite with 4 business apps and 4 prompts, on 8.8 of 50 free credits. The dashboard maths was exactly right, and the first click on the live order form did nothing. Both made the review.
Brands get my findings before I post, so they can fix issues first.
Tools I cover: AI agents, agent tools and APIs, automation, AI video and image generation.
Building an AI tool? Paid collabs only, DMs open.
CLEMENT AI retweeted
I spent $0 and 8.8 of 50 free credits testing @zite.
4 business apps, 4 prompts, 0 errors. The dashboard math was exactly right. Then my first click on the live order form did nothing.
Access: free account with Google sign-in. 50 credits a month and 10 messages a day. No card, no bonus or referral credits.
What I tested
Easy: to-do app for a 3 person bakery team. About 2 min 15 s, about 2 credits
Typical: order tracker from my own CSV of 8 orders. About 2 min, about 2 credits
Hard: public order form, revenue dashboard and staff-only login in one prompt. About 2 min, 3.6 credits
Extra: change color and title by chat (40 s, 1.2 credits), then publish (free)
What it does well
→ Every build worked first time, no errors
→ All 8 CSV orders imported with the right totals
→ Dashboard showed $168.35, then $181.35 after a form order. Both correct
What it does badly
→ First "Place order" click on the live form did nothing
→ The order form lets customers type their own price
→ Adds things I did not ask for: sample tasks, a Delete option, a new subtitle
→ Made a second app instead of a page
→ Free plan: 10 messages a day, "Made with Zite" badge
→ Terms are Fillout's from 2022 with nothing on AI. Training opt-out only on Business
Pricing as listed
Free $0: 50 credits/mo, 10 messages a day
Pro $25/mo ($20/mo yearly): remove branding, custom domain
Business $89/mo ($75/mo yearly): AI training opt-out by default
Scores
Prompt to app 8/10
CSV to order tracker 9/10
Form, dashboard and login 8/10
Edit and publish 7/10
Free tier value 6/10
Privacy and terms 5/10
Speed and UI 8/10
Verdict 7.3/10
Use it for small internal business apps on your own data.
Skip it if you need no branding or clear AI privacy terms on the free plan.
What I would fix first
→ Clear AI terms and a training opt-out on every plan
→ Fix the dead first click on form submit
→ Lock prices on order forms by default
→ Ask before adding things I did not request
Tip: upload your own spreadsheet in the first prompt. It saves messages, and you only get 10 a day on Free.
CLEMENT AI retweeted
A freelance designer in Joburg finished three jobs this month. He has been paid for one.
He is good at his work. Chasing money is the part he hates, so he puts it off, and the invoices get older.
Three things keep him waiting.
One: his invoices live in his email, a spreadsheet and his head. He only notices one is late when rent is due.
Two: he does not want to sound rude, so his reminders are too soft, or he sends none at all.
Three: he treats a client who is 3 days late the same as one who is 30 days late.
The fix is an AI agent that keeps track and writes the reminder for him.
Here is what he would type into @CreaoAI, in plain words:
"Build me an invoice follow-up assistant. I will give you my client list with the invoice number, amount, due date and whether it has been paid. When I ask, tell me which invoices are overdue and by how many days. For each one, draft a reminder in a polite, professional tone: friendly at 7 days late, firmer at 14 days, and a final notice at 30 days. Use the client's name and the invoice number. Never send anything yourself. I will check every message before it goes."
That message builds the agent. No coding.
He still decides who gets which message and when. He just never has to start from a blank email again.
We are building and testing this one now. Real results go up on fix desk once we have them.
Freelancers, how late do your clients usually pay? Reply with what you do and your country.
This is the Joburg freelancer from our pinned post. Every story in that post is getting its own fix. Freelancers, reply with your work and the reminder you find hardest to send, and I will tailor the prompt to it.
I spent $0 and 8.8 of 50 free credits testing @zite.
4 business apps, 4 prompts, 0 errors. The dashboard math was exactly right. Then my first click on the live order form did nothing.
Access: free account with Google sign-in. 50 credits a month and 10 messages a day. No card, no bonus or referral credits.
What I tested
Easy: to-do app for a 3 person bakery team. About 2 min 15 s, about 2 credits
Typical: order tracker from my own CSV of 8 orders. About 2 min, about 2 credits
Hard: public order form, revenue dashboard and staff-only login in one prompt. About 2 min, 3.6 credits
Extra: change color and title by chat (40 s, 1.2 credits), then publish (free)
What it does well
→ Every build worked first time, no errors
→ All 8 CSV orders imported with the right totals
→ Dashboard showed $168.35, then $181.35 after a form order. Both correct
What it does badly
→ First "Place order" click on the live form did nothing
→ The order form lets customers type their own price
→ Adds things I did not ask for: sample tasks, a Delete option, a new subtitle
→ Made a second app instead of a page
→ Free plan: 10 messages a day, "Made with Zite" badge
→ Terms are Fillout's from 2022 with nothing on AI. Training opt-out only on Business
Pricing as listed
Free $0: 50 credits/mo, 10 messages a day
Pro $25/mo ($20/mo yearly): remove branding, custom domain
Business $89/mo ($75/mo yearly): AI training opt-out by default
Scores
Prompt to app 8/10
CSV to order tracker 9/10
Form, dashboard and login 8/10
Edit and publish 7/10
Free tier value 6/10
Privacy and terms 5/10
Speed and UI 8/10
Verdict 7.3/10
Use it for small internal business apps on your own data.
Skip it if you need no branding or clear AI privacy terms on the free plan.
What I would fix first
→ Clear AI terms and a training opt-out on every plan
→ Fix the dead first click on form submit
→ Lock prices on order forms by default
→ Ask before adding things I did not request
Tip: upload your own spreadsheet in the first prompt. It saves messages, and you only get 10 a day on Free.
CLEMENT AI retweeted
I spent $0 testing @Curriculoai, a free AI hiring tool that scores resumes against a job.
It ranked my 5 test resumes in exactly the right order. Then it told me my remote job "requires on-site work", and its "Enhance with AI" button deleted my salary line.
Here's what works and what doesn't.
Disclosure: free Starter plan, signed up with Google, no card. No bonus or referral credits. $0.
What I tested
Easy: create a job from my own post. About 1 min 20 s, $0
Typical: upload 5 made-up PDF resumes and check scores. First scores in about 2 min, all 5 in about 5 min, $0
Hard: catch a keyword stuffer, score a Spanish resume fairly, answer a question about candidates. $0
What it does well
→ Right order: 92, 90, 65, 32, 2, same as my own ranking
→ Caught the keyword stuffer (32) and scored the Spanish resume fairly (90)
→ Free plan is real: 1,000 candidates, no card
What it does badly
→ Says my remote job requires on-site work, and calls it remote-first in the same summary
→ "Enhance with AI" dropped my salary and added Git and CI/CD
→ Added CI/CD as a qualification, then marked everyone down for it
→ "Screening invite sent" on 2 people with no email. I sent nothing
→ No CSV import for applicants
→ Toolbar archive button did nothing, twice
→ List didn't refresh: one score only showed after a reload
Pricing as listed
Starter: $0, free forever, up to 1,000 candidates
Pro: $50/mo, 10,000 candidates
Scores
Job setup 6/10
Resume scoring 7/10
Bulk CSV import 3/10
Pipeline and AI extras 4/10
Verdict 5.5/10
Use it for a free first-pass ranking of a pile of PDF resumes for one role.
Skip it if you need CSV imports, reliable invite emails, or remote-aware screening.
What I'd fix first
→ Read remote vs on-site from the job itself
→ Only show "invite sent" when an email went out
→ Keep salary and requirements in AI rewrites
→ Add CSV import for applicants
Tip: write your own job post, skip the AI rewrite, and delete any qualification it adds that you didn't ask for. It scores every candidate against them.
Bookmark this if you hire with AI tools.
I spent $0 testing @Curriculoai, a free AI hiring tool that scores resumes against a job.
It ranked my 5 test resumes in exactly the right order. Then it told me my remote job "requires on-site work", and its "Enhance with AI" button deleted my salary line.
Here's what works and what doesn't.
Disclosure: free Starter plan, signed up with Google, no card. No bonus or referral credits. $0.
What I tested
Easy: create a job from my own post. About 1 min 20 s, $0
Typical: upload 5 made-up PDF resumes and check scores. First scores in about 2 min, all 5 in about 5 min, $0
Hard: catch a keyword stuffer, score a Spanish resume fairly, answer a question about candidates. $0
What it does well
→ Right order: 92, 90, 65, 32, 2, same as my own ranking
→ Caught the keyword stuffer (32) and scored the Spanish resume fairly (90)
→ Free plan is real: 1,000 candidates, no card
What it does badly
→ Says my remote job requires on-site work, and calls it remote-first in the same summary
→ "Enhance with AI" dropped my salary and added Git and CI/CD
→ Added CI/CD as a qualification, then marked everyone down for it
→ "Screening invite sent" on 2 people with no email. I sent nothing
→ No CSV import for applicants
→ Toolbar archive button did nothing, twice
→ List didn't refresh: one score only showed after a reload
Pricing as listed
Starter: $0, free forever, up to 1,000 candidates
Pro: $50/mo, 10,000 candidates
Scores
Job setup 6/10
Resume scoring 7/10
Bulk CSV import 3/10
Pipeline and AI extras 4/10
Verdict 5.5/10
Use it for a free first-pass ranking of a pile of PDF resumes for one role.
Skip it if you need CSV imports, reliable invite emails, or remote-aware screening.
What I'd fix first
→ Read remote vs on-site from the job itself
→ Only show "invite sent" when an email went out
→ Keep salary and requirements in AI rewrites
→ Add CSV import for applicants
Tip: write your own job post, skip the AI rewrite, and delete any qualification it adds that you didn't ask for. It scores every candidate against them.
Bookmark this if you hire with AI tools.
CLEMENT AI retweeted
A caterer in Kumasi gets the same message every weekend: "How much for a party of 150?"
She cannot answer straight away. Every quote means her notebook, a calculator and a call to the meat seller.
By the time she replies, the customer has booked someone faster.
Three things slow her down.
One: her prices live in her head and a notebook, so every quote starts from zero.
Two: customers leave things out. No date, no venue, no guest count. She has to ask before she can price anything.
Three: she writes quotes late at night after cooking. Small costs like transport and extra serving staff get forgotten, and she pays for them herself.
The fix is an AI agent that turns an enquiry into a quote.
Here is what she would type into @CreaoAI, in plain words:
"Build me a catering quote assistant. Here is my price sheet: [paste it]. When I paste a customer enquiry, check it has the date, venue, number of guests and menu choice. If anything is missing, write a short polite message asking for it. If everything is there, work out a full quote including food, serving staff and transport, show the sums line by line, and write it as a friendly message I can send. Never use a price that is not on my sheet."
That message builds the agent. No coding.
She pastes the enquiry, checks the sums and sends the quote. She still sets her prices and chooses her jobs.
We are building and testing this one now. Real results go up on fix desk once we have them.
Fix desk, day 3.
Caterers and event vendors, what does a customer always forget to tell you? Reply with your business and your country.
Not a caterer? The same agent works for any business that sends quotes: tailors, decorators, printers, movers. Reply with what you sell and I will adapt the prompt for you.
CLEMENT AI retweeted
24 hours ago this YouTube channel didn't exist. Its first video now has 995 views.
Day 1 of taking an AI-built faceless channel from 0 to 100K subscribers.
@CreaoAI built the whole thing: the name, the character, the idea, the script, every scene, the voiceover, the captions and the upload. I didn't open an editor once.
The first Short: "Noodles at 11pm, off the good plate."
Day 1 numbers, no filter:
→ 990 views
→ 10 likes
→ 1 comment
→ 4 subscribers
Small numbers, but real ones, from a channel with zero followers and no promotion behind it.
What I'm watching next:
→ Does the character stay the same in episode two and three? This is the real test. My last channel drifted after a few days.
→ Do views turn into subscribers? 990 views and 4 subs tells me the video gets watched, but people don't have a reason to follow yet.
That second one is the next fix. I'll show you what I change and whether it works.
99,996 subscribers to go. Subscribe to the channel to watch it grow, and follow me here for the honest numbers every step of the way.
I didn't touch CapCut once. @CreaoAI built my faceless YouTube channel from scratch, and the first video is already live.
Quick context. My last channel ran on Grok Bot, and the character kept drifting, even after I simplified the task. I needed to know whether my setup was the problem or the tool.
So I ran a fair test. Brand-new channel. One agent on CREAO, given four rules:
→ Continuity rule: every scene follows on from the last
→ Natural realism rule: she has to look and move like a real person
→ Consistency rule: her face, body and look never change
→ Negative rule: a clear list of what she must never do or become
Then I let it build everything:
→ CREAO picked the channel name
→ CREAO created the character and her bible sheet
→ CREAO wrote the idea
→ CREAO wrote the script
→ CREAO created and directed the scenes
→ CREAO made the video clip by clip and merged it, with no CapCut
→ CREAO added the voiceover
→ CREAO added the captions
→ CREAO posted it on YouTube
It's live now. The link is in the first reply, so go and judge it yourself.
I said last time I'd split the work into three agents. I'm testing one first, on purpose. If one agent on CREAO holds the character, the tool was the problem. If it drifts, I'll split it and find out exactly what broke.
We're taking this channel from zero to 100K subscribers, and I'll post what works and what breaks, so you don't make the same mistakes.
Subscribe to the channel to watch it grow, and follow me here so you don't miss a single update.
