Fly Labs
Ask FlyBot

Libretto – Making AI browser automations deterministic

Libretto (https://libretto.sh) is a Skill+CLI that makes it easy for your coding agent to generate deterministic browser automations and debug existing ones. Key shift is going from “give an agent a prompt at runtime and hope it figures things out” to: “Use coding agents to generate real scripts you can inspect, run, and debug”.Here’s a demo: https://www.youtube.com/watch?v=0cDpIntmHAM. Docs start at https://libretto.sh/docs/get-started/introduction.We spent a year building and maintaining browser automations for EHR and payer portal integrations at our healthcare startup. Building these automations and debugging failed ones was incredibly time-consuming.There’s lots of tools that use runtime AI like Browseruse and Stagehand which we tried, but (1) they’re reliant on custom DOM parsing that's unreliable on older and complicated websites (including all of healthcare). Using a website’s internal network calls is faster and more reliable when possible. (2) They can be expensive since they rely on lots of AI calls and for workflows with complicated logic you can’t always rely on caching actions to make sure it will work. (3) They’re at runtime so it’s not interpretable what the agent is going to do. You kind of hope you prompted it correctly to do the right thing, but legacy workflows are often unintuitive and inconsistent across sites so you can’t trust an agent to just figure it out at runtime. (4) They don’t really help you generate new automations or help you debug automation failures.We wanted a way to reliably generate and maintain browser automations in messy, high-stakes environments, without relying on fragile runtime agents.Libretto is different because instead of runtime agents it uses “development-time AI”: scripts are generated ahead of time as actual code you can read and control, not opaque agent behavior at runtime. Instead of a black box, you own the code and can inspect, modify, version, and debug

Hacker NewsToolAISource
0Sign in to voteCopy link

FL score

68

out of 100

Verdict

VALIDATE

high confidence

Competition

14

competitors found, emerging market, funded players

Trend

No signal yet

Libretto aims to make AI browser automations deterministic and debuggable by generating inspectable code at development-time, targeting complex, high-stakes enterprise environments.

The pain

Existing AI browser automation tools are unreliable, expensive, and opaque due to runtime agent unpredictability, making debugging and maintenance incredibly time-consuming, especially for complex or legacy websites in high-stakes industries like healthcare.

The gap

There's a significant gap between opaque, non-deterministic runtime AI agents and the heavy manual coding required for traditional frameworks. A solution that uses AI to *generate inspectable, deterministic code* for browser automations, providing control and debuggability, is currently unaddressed, particularly for critical enterprise use cases.

Build angle

Build a CLI/skill that, given a high-level goal, leverages AI to generate well-structured, Playwright-based code for browser automations, with built-in debugging capabilities. Focus initially on a narrow niche like healthcare EHR/payer portal interactions, emphasizing deterministic outcomes and inspectable code.

Strengths

  • Addresses a severe, costly, and well-understood pain point (unreliable/undebuggable automations).
  • Unique 'development-time AI' approach generates inspectable, deterministic code, providing a strong differentiator.
  • Founder's direct experience provides strong problem insight and credibility.
  • High willingness to pay for reliability and reduced operational risk in high-stakes environments.
  • Growing market for AI and automation, with increasing need for robust solutions.

Risks

  • High technical complexity to build and maintain the core AI code generation system as a solo founder.
  • Difficulty in achieving true determinism and broad applicability across diverse, dynamic websites.
  • Market is crowded with well-funded incumbents, requiring a very precise niche and distribution strategy.
  • Audience reach for enterprise clients can be challenging for a micro-SaaS.
  • Longer time-to-market for a robust, truly valuable MVP.
  • Risk of the generated code still requiring significant manual tweaking or being too brittle for rapid UI changes.

Next steps

  1. Conduct in-depth interviews with developers and automation engineers in target industries (beyond just healthcare) to further validate the 'development-time AI' approach and specific features needed.
  2. Build a proof-of-concept that demonstrates the deterministic code generation for 2-3 specific, complex real-world workflows that existing runtime AI tools fail at.
  3. Gather feedback on the generated code quality, debuggability, and ease of modification.
  4. Investigate distribution channels beyond traditional developer communities, considering direct outreach or industry-specific partnerships.
  5. Scope a minimal viable product (MVP) with extremely tight focus, perhaps on a single, critical automation task for a specific type of website, to prove out the core value proposition rapidly.

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “Libretto – Making AI browser automations deterministic”.

Open FlyBot