muchael

6 ms·

1.
by muchael 2mo ago
Appreciate it! Yea it's much cheaper and works better with how models are already trained on playwright to interact with browser
2.
by muchael 2mo ago
We set up the GitHub integration and authentication for you so all you have to do is add one line into your code to enable opening the PRs It’s all open source so you can just take the skill + CLI if you want to. But the GitHub integration
3.
Show HN: Libretto PR agents – Automatically fix failing playwright scripts(libretto.sh)
22 points by muchael 2mo ago|5 comments
4.
by muchael 5mo ago
Our philosophy with building automation scripts is to use AI in a limited/intentional way. From our experience writing RPA scripts, there are rarely website changes on super old portals. Also using internal network requests is more rel
5.
by muchael 5mo ago
Thanks! I think the right solution is 100% a mixture. Currently thinking it should be mostly deterministic with some intentional/limited usage of runtime AI. And then AI debugging tooling on top of that
6.
by muchael 5mo ago
For our scripts running in prod we handle this in 2 ways: - We use runtime agents in very specific places. For example on Availity they frequently have popups right after you login, so if there's a failure right after signup we spin up
7.
by muchael 5mo ago
Answered similar question above responding to someone's MCP sampling comment. Spinning up a separate agent in the CLI was our initial approach but we switched to snapshot via API because of speed and reliability. We can update the conf
8.
by muchael 5mo ago
Not totally sure I understand, but if you're talking about the snapshot command which requires an API key we initially had it spinning up a tmux session to analyze the snapshot instead of using the API. But we switched it to use the AP
9.
by muchael 5mo ago
This is a great flag and something we want to spend more time experimenting with as we continue to build out the repo. Right now we kind of have a mixture of the 2 approaches, but there's a large room for improvement. - When libretto p
10.
by muchael 5mo ago
Agreed! One thing that we felt was missing from the existing MCP tools was user recording. For old and shitty healthcare websites it's easier to just show the workflow than explain it The playwright codegen tool exists, but the script
11.
by muchael 5mo ago
There are a couple ways to handle JS components rendered at runtime: - Libretto prefers network requests over DOM interaction when possible, so this will circumvent a lot of complex JS rendering issues - When you do need the DOM, playwright
12.
by muchael 5mo ago
Haha that's wild, let me know if you run into any issues with it!
13.
by muchael 5mo ago
Super cool! Please let me know how it goes. Since agents are so good at writing code, we think letting the agent rewrite/test the code on failure is better than just using a prompt at runtime
14.
by muchael 5mo ago
Right now libretto only captures HTTP requests, which the coding agent can use to determine how to perform the automation. For more complex cases where libretto can't validate that the network approach would produce the right data (lik
15.
by muchael 5mo ago
Lol sorry for the misleading click. We named it libretto after the term in theater, inspired by Playwright. No retro gaming here, just browser automation!
16.
by muchael 5mo ago
It's a good callout. We have a BAA + ZDR with Anthropic and OpenAI, and if you want to use libretto for healthcare use cases having a BAA is essential. Was using Codex in the demo, and we've seen that both Claude and Codex work pr
17.
by muchael 5mo ago
At its core, libretto generates, validates, and helps with debugging RPA scripts. As far as I understand tools like playwright CLI are more focused on letting your agent use playwright to perform one-off automations. The implementation is a
18.
Show HN: Libretto – Making AI browser automations deterministic(github.com)
134 points by muchael 5mo ago|55 comments
19.
Show HN: Job Offer Compensation Calculator(share.streamlit.io)
2 points by muchael 3y ago|1 comments