Frances and I really appreciated the feedback from our first launch. Today we’re launching Promptless 1.0, which addresses our biggest learnings from the last 12 months.
I also made it way easier to try it out. You can tag @promptless on any open-source Github PR or Issue with a doc update request, and Promptless will create a fork and open a PR for your docs to help. Feel free to use our own docs as a playground: https://github.com/Promptless/docs/issues
Or, you can sign up at https://promptless.ai to get free access for your own docs for the next 30 days. Here's a demo video: https://youtu.be/IWwimHCEY7Y
For me, the coolest part of the last year has been seeing how users got creative with Promptless. One user has Promptless listening in to all their Slack Connect channels, so whenever they answer a customer question, Promptless figures out if their docs should be updated and drafts an update if so. Another user has Promptless processing every customer meeting transcript and updating their internal docs after each meeting: customer dashboards, feature request pages, etc.
Some of the biggest things that are new with version 1.0:
- Automatically updating screenshots: this was by far our most requested feature. The need here was always clear. People would exclude screenshots from docs because they’d get stale quickly, even though they knew screenshots would be helpful to users. A year ago, we just couldn't ship a good enough solution, but given how much LLMs' visual grounding has improved in the last year, now we've got something we're proud of.
- Slop-free writing: The most common critique on early Promptless suggestions was that even though they were accurate, they could sound generic or verbose, or might just reek of AI slop. Promptless 1.0 is 3.5x better at this (measured by voice-alignment compared to what users actually published), through a combination of fine-tuned models, sub-agents, and alignment on user-defined preferences.
- Open-source program: We're especially proud of this—Promptless is now free for CNCF/Linux Foundation projects (reach out if you’re a maintainer!). You can take a look at how Promptless is supporting Vitess (a CNCF-graduated project) with their docs here: https://github.com/vitessio/website/commits
Check it out and let us know if you have any questions, feedback, or criticism!
duskwuff•1h ago
By running the software and getting it into an equivalent state to the original screenshot? Impressive if true.
On the other hand, if you're just "updating" the screenshot by making the AI model imagine what it would look like with some code changes - that's worse than useless; it's likely to end up generating screenshots that won't match any released version of your software at all.
prithvi2206•39m ago
Here's an example from our own docs, where you can see the captured screenshots if you scroll down to the PNGs that Promptless added https://github.com/Promptless/docs/pull/233/changes
Getting the browser navigation to work was predictably hard, but the other thing that's very hard to get right is timing when Promptless takes the screenshots. Some teams want Promptless capturing screenshots on a staging environment, and sometimes it's on prod, and different teams have different release cycles and feature flagging processes.
For that reason, we still have to do a bit of manual set-up for each user that wants screenshot updating.
Agreed though that just asking nano banana to predict a new screenshot is negative value.