Your project has 48 facts, 31 of them implemented: code-backed, verified by command. 12 are specs your agent is working through. 5 are rough drafts you'll refine later. You know all of this because you ran facts check.
I'm a bit tired of spending my time working on large specifications and making agent implement them deviating somewhere in the middle forcing rework loops, so I'm switching to a more basic primitive - facts.
Facts are just tiny assertions about the expected project state. They are much faster to read through compared to the specs. Faster to validate. Facts are atomic - your agent can fully autonomously verify if your project is still in the state you want it to be. Facts can be codified to run like a linter for your project quality. Facts can be refined from high-level ideas to the ground truth about the project. Facts can be mined from an existing codebase and cross-verified for consistency.
Facts are the antidote to fluffy specs.
"48 facts, 31 implemented, 12 in progress, 5 drafts" — that's a perfect framing of what most engineering specs lose the moment they leave Notion. We have the same fluff problem in financial modeling. "DSCR target 1.30x, sculpted debt, partnership flip after Year 6" sounds rigorous on a slide and turns out to be three different things to three reviewers. The fix is the same: facts that are code-backed and verifiable from the model itself, not prose that ages out. That's the discipline I tried to bake into my project finance and valuation templates on Eloquens — testable assumptions, not narrative. Curious how Facts handles spec drift after a fact is verified — do you flag re-verification when the underlying code changes, or only on commit?
Hi PH, I'm absolutely blown away by the response to this small project, thank you!
As a token of my gratitude, I want to share the workflow I used for the promo video:
1. I start all such projects with a definition of design/visual language. This time it was around PS1-era aesthetic. Initial designs were nothing like the final, but it clicked eventually when I suggested to use Spyro as a reference.
Tip: there are websites to preview color palettes in realistic settings via URL, this way your agent can share ideas with you very quickly.
2. In my lifeos, I maintain a few content templates, one of them is for a motion graphics script creation, I use it to create a first draft of the script.
Tip: I usually start and center the script around a single central idea or state the viewer should receive. After initial version is done, I assemble initial video, just to see where it lands.
3. Adjusting the visual style and look'n'feel until it looks as intended. In this instance, for example, I've tweaked leading fonts, the specific look'n'feel of rendering, and added two different color palettes for text and visuals so that they don't overlap. After this step the visual part is locked and I proceed iterating on the content.
4. This is just going through all scenes one-by-one and verifying that they make sense. One easy tip is to always ask agent to redo the scenes considering viewer has zero context about the project or its purpose in mind as well as make every scene ~5s longer than you think it needs to be.
As a bonus, with the visual/desing language codified - you can later proceed creating the content for the same project without loosing the identity, all in agentic-friendly way :)
Thanks again!
About Facts on Product Hunt
“The antidote to fluffy specs.”
Facts launched on Product Hunt on May 5th, 2026 and earned 72 upvotes and 5 comments, placing #28 on the daily leaderboard. Your project has 48 facts, 31 of them implemented: code-backed, verified by command. 12 are specs your agent is working through. 5 are rough drafts you'll refine later. You know all of this because you ran facts check.
Facts was featured in Developer Tools (512k followers), Artificial Intelligence (467.7k followers), GitHub (41.2k followers) and Vibe coding (438 followers) on Product Hunt. Together, these topics include over 180.2k products, making this a competitive space to launch in.
Who hunted Facts?
Facts was hunted by Ivan Charapanau. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Facts stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Ivan Charapanau