danoandco

6 ms·

1.
The rate trap: how one architecture decision kills flexibility(github.com)
1 points by danoandco 5mo ago|0 comments
2.
by danoandco 5mo ago
Definitely and Twill is for SWE delegation first, not so much the “general agent on my machine.”
3.
by danoandco 5mo ago
It's a crowded market. On the CLI-agnostic cloud agent positioning, there are only startups so far. Only incumbent is Github Agents as you mentioned in another thread.
4.
by danoandco 5mo ago
Yes, broadly. The main structural difference is that we’re agent-agnostic, so we can combine lab-native CLIs in one workflow. GitHub will likely struggle there because they have direct partnerships with Anthropic and OpenAI. On the features
5.
by danoandco 5mo ago
On gh-aw: it looks solid for the event-driven automation shape (triage, docs sync, CI fix). We're after a slightly different shape: interactive back-and-forth, steering from Slack or Linear, persistent sandboxes with a booted dev serve
6.
by danoandco 5mo ago
Mmh this works on my end. Sending you an email. Ty
7.
by danoandco 5mo ago
On computer use: Yes. Sandboxes come with a computer-use CLI for driving Linux GUI apps via X11. On triggers: Cron, GitHub (PRs, issues, @twill mentions in review comments), Slack, Linear, Notion, Asana webhooks, plus CLI and web. Our PR-co
8.
by danoandco 5mo ago
Awesome! Thanks for trying it.
9.
by danoandco 5mo ago
Jules is similar to Twill with the following differences: - Twill is CLI-agnostic, meaning you can use Claude Code, Codex or Gemini. Jules only works with Gemini. - We focus on the delegation experience: Twill has native integrations with y
10.
by danoandco 5mo ago
On the Twill web app, you can run the same task across different agents and multiple attempts (each in its own sandbox). Then you pick the best result. This is super handy for UI work where you can open the live preview for each attempt and
11.
by danoandco 5mo ago
Similar but reusing lab-native CLIs like Claude Code or Codex, which they perform RL on. And so in the long-run, we believe this approach wins over custom harnesses.
12.
by danoandco 5mo ago
We’re focused on SWE use cases. Code is nice because there’s already a built-in verification loop: diffs, tests, CI, review, rollback. But you do quickly get to a state where the agent needs to make a risky action (db migration, or an infra
13.
by danoandco 5mo ago
Totally right on the compile time. CIs have the same bottleneck, and the ecosystem is working on fixing this (faster cpus, better caching) in both coding agents and CI to improve overall velocity
14.
by danoandco 5mo ago
For a solo dev running one task at a time, a beefy desktop overnight is totally viable. We see a lot of this with the Mac Mini hype Cloud starts to matter when you want to (a) run a swarm of agents on multiple independent tasks in parallel,
15.
by danoandco 5mo ago
Yes, this is the pass@k metric from code generation research. Found the relevant paper Evaluating Large Language Models Trained on Code (Chen et al., 2021) which introduced the metric.
16.
by danoandco 5mo ago
Claude managed agents is a general-purpose hosted runtime for Claude. While Twill focuses on SWE tasks. And so the SWE workflow is pre-built (research, planning, verification, PR, proof of work). Twill is also agnostic to the agent, so you
17.
Launch HN: Twill.ai (YC S25) – Delegate to cloud agents, get back PRs(twill.ai)
77 points by danoandco 5mo ago|95 comments
18.
Clone any web app in minutes(twill.ai)
4 points by danoandco 6mo ago|0 comments
19.
by danoandco 6mo ago
Thanks for running it and the feedback! For the ADR vs AGENTS: CLIs usually load the AGENTS.md with a tag saying: "this context may or may not be relevant to your tasks. You should not respond to this context unless it is highly releva
20.
Agent skills for desktop automation and video recording(github.com)
3 points by danoandco 6mo ago|0 comments
21.
by danoandco 6mo ago
true, i think the key thing is explaining somewhere in the repo "why" something was done. like the rationale for choosing X over Y service for instance. maybe this record is just the git log, and the agent just needs to access the
22.
by danoandco 6mo ago
OpenAI published an article and demo for scoring how well AI agents can work in a codebase ( https://openai.com/index/harness-engineering/ , https://www.youtube.com/watch?v=rhsSqr0jdFw ). We turned i
23.
Show HN: Score your GitHub repo for AI coding agents(twill.ai)
7 points by danoandco 6mo ago|5 comments
24.
by danoandco 1y ago
reminds me of the "memory for agents is a moat" but for software memory for software
25.
Show HN: Chainlit Now Supports Anthropic's Model Context Protocol (MCP)(docs.chainlit.io)
1 points by danoandco 1y ago|0 comments
26.
Embed your own Python AI Agent on your software with Chainlit(twitter.com)
1 points by danoandco 2y ago|1 comments
27.
by danoandco 2y ago
Introducing Chainlit Copilot mode You can now embed a Chainlit app in an existing software - Function Calling: your Copilot can even take actions on the website - Widget UI & UX customisation - CORS & Authentication to manage who ca