Every explainer video I tried to make died in the timeline keyframes, re-timing, redo. So I built manic: you write a few plain lines (draw this, move that, flash it green) and it renders a clean animated video. Works for math, algorithms, data structures, and now 3D.

It's for people who don't want to open After Effects teachers, devs writing docs, anyone explaining something visual.
Manic https://8gwifi.org/manic/
Manic Docs http://8gwifi.org/manic/docs

Hey everyone

I'm Shiva, solo dev, and I just rebuilt something I first put out a few months ago: FreeViralKit a free AI toolkit for YouTube creators. Title generator, tag generator, description optimizer, hook writer, niche research, script outlines. No paywall on the core tools.

I built it because I run a small horror storytelling channel myself and got tired of guessing titles/tags at 1am with zero budget for TubeBuddy or VidIQ.

First launch flopped hard 3 upvotes, mostly bot traffic. I've since rewritten the copy (it read way too corporate before) and fleshed out the tool pages properly.

Hey everyone and hi product hunt,

I want to share what I ve been working on. The current trend in AI agent security is using another LLM as a "guardrail" to check if a prompt is malicious. I think this is fundamentally broken it's slow, probabilistic, and gets bypassed by indirect prompt injections.

AI agents are the ultimate "Confused Deputy." They hold high-privilege API keys, but their reasoning can be hijacked by untrusted data (like a Jira ticket or an email). The real damage doesn't happen at the prompt layer; it happens at execution.

So I built Relay a runtime security proxy that sits between the agent's decision and the actual tool execution.

Hi Product Hunt I m building AI Voice Wallet, a Telegram-based expense tracker for people who prefer speaking or typing naturally instead of filling out a form. A message like spent 18 AED on lunch and 10 on a cab becomes separate structured records, with summaries and budgets afterward. It does not connect to bank accounts; reports reflect what users record. I d value feedback on whether voice-first entry is useful enough to replace a lightweight manual tracker. Try it: https://t.me/voice_wallet_bot

Problem

Dropshippers and e-commerce sellers spend hours manually researching competitors checking their ads, estimating revenue, tracking SEO, and collecting winning creatives. It's time-consuming, fragmented, and takes away from actually growing their business.

Most portfolio tools I found for actors and artists had one of three problems: they looked dated, cost more than an early-career artist wants to spend, or made every content update depend on a developer.

So I built ArtistPass: a polished public portfolio plus a browser-based Admin for credits, gallery, showreel and contact details.

I nearly kept expanding it into a multi-template SaaS. Instead, I stopped at a useful open-source release and put it in the open to see what real demand appears.

Live demo: https://artistpass.vercel.app

Create Premium Short Videos With AI.
No Editing Team Required.

Zynos AI generates publishing-ready TikToks, Reels, and YouTube Shorts with talking avatars, faceless b-roll, and viral hooks. Professional output. Zero generic AI slop.

Hey Product Hunt community!

If you've worked with proxies at scale, you know the pain: frequent IP bans killing your scraping jobs, unstable connection speeds derailing automation workflows, opaque pricing that balloons unexpectedly, and integration docs that make even simple setups take hours. We've been there too.

That's why we built Pxyedge a rotating residential proxy platform designed for developers and growth teams who need reliability without the headaches.

Here's what we're bringing to the table:

I m trying to understand how builders are handling this.
AI agents are moving from generate text to take actions :
- run shell commands
- call APIs
- modify GitHub repos
- access databases
- trigger deploys
- send webhooks/emails
- use MCP tools
Prompt injection feels much more serious once the agent can actually execute something.
My current thinking is that teams need a checkpoint before execution:
agent wants to call a tool
policy checks the tool + args
action is allowed, blocked, or logged
agent continues or stops
I m building a small tool around this called Relay, but I m mostly looking for feedback from people building agents.
What would you block by default?
My current list:
- destructive shell commands
- repo deletion / force push
- database drop / truncate
- production deploys
- secret access
- outbound calls to unknown domains
- emails/webhooks with sensitive data
Curious how others are solving this today:
- allowlists?
- denylists?
- human approval?
- sandboxing?
- audit logs?
- MCP gateway?
- custom wrappers?
Relay is here if useful:
https://relay-security-lemon.ver...

After trying dozens of APIs for SEO, SERP data, scraping, AI, and social media, I got tired of juggling different providers, authentication methods, and pricing models.

That's why I've been using YepAPI.

It brings 130+ endpoints together under a single API key, covering things like:

Google SERP & keyword data
Backlink and domain analysis
Web scraping with JavaScript rendering
AI models through one OpenAI-compatible endpoint
YouTube, TikTok, and Instagram data