Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
YaraDori
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
1.
▲
by
YaraDori
7mo ago
This is wild — using iPhone Mirroring as the “device API” is a clever hack. Two reliability questions: 1) How do you handle UI drift / A-B layouts / dynamic text (e.g., different font sizes, localization)? Do you anchor on vision+
2.
▲
by
YaraDori
7mo ago
Strongly agree with the problem framing. In practice I’ve seen two failure modes: 1) “compression amnesia” inside a single long session (decisions disappear silently) 2) “tool silo” memory (each agent/tool has its own scratchpad, nothi
3.
▲
by
YaraDori
7mo ago
This is a neat angle — “AI session doesn’t end after the first draft” is exactly the missing piece in most form builders. Two questions from the automation/agent side: 1) Do you store the user’s edits as structured deltas (so the syste
4.
▲
by
YaraDori
7mo ago
Cool idea — “book as rubric” is a nice way to avoid vague self-critiques. One thing I’ve found helps keep agent review loops from getting shallow is to separate levels of critique: 1) a fast “lint” pass (format, obvious bugs, missing test
5.
▲
by
YaraDori
7mo ago
I love the “mandatory approval + audit log” stance — feels like the only sane way to run agents near real accounts. Curious how you think about the reliability layer for UI work: - do you store any per-step evidence (DOM snapshot/scree
6.
▲
by
YaraDori
8mo ago
Hey HN, I built SkillForge because I got tired of manually writing agent instruction files. The idea: record yourself doing a task on your computer, and AI extracts the workflow into a structured SKILL.md file that agent frameworks (Claude
7.
▲
Show HN: SkillForge – Turn screen recordings into agent skill files
(skillforge.expert)
4 points
by
YaraDori
8mo ago
|
1 comments
8.
▲
by
YaraDori
8mo ago
Hey HN! Creator here. OpenClaw is an open-source personal AI agent that runs 24/7 on your own hardware. What makes it different from ChatGPT/Claude: - It connects to your actual messaging apps (WhatsApp, Telegram, Slack, Discord,
9.
▲
Show HN: OpenClaw – Open-source personal AI agent that lives on your machine
(github.com)
2 points
by
YaraDori
8mo ago
|
3 comments
10.
▲
by
YaraDori
8mo ago
Creator here. SkillForge watches what you do on screen and turns it into structured SKILL.md files that AI coding agents (Claude Code, Codex, etc.) can use. Instead of writing long instruction docs, you just demonstrate the workflow and Ski
11.
▲
Show HN: SkillForge – Turn screen recordings into AI agent skills (SKILL.md)
(skillforge.expert)
2 points
by
YaraDori
8mo ago
|
1 comments
12.
▲
by
YaraDori
8mo ago
This hits a real pain point. Writing SKILL.md files by hand for agent frameworks sounds simple but takes forever - you have to think through every edge case and decision branch. The screen recording approach is smart because it captures imp
13.
▲
by
YaraDori
8mo ago
The most valuable investment in 2026 is turning your daily personal workflow, which usually spread across multiple apps, into a fully automated agentic skill that allows your personal agent to take over on your local computer. That’s exactl
14.
▲
by
YaraDori
8mo ago
Self-hosting agents on spare machines is a great model — the privacy angle especially resonates for enterprise users. To your question about what tasks to automate: one bottleneck I keep hitting is the "teaching" step. Even with a
15.
▲
by
YaraDori
8mo ago
Cool approach! The plain-English-to-automation idea resonates with something I've been exploring too. One thing I've noticed is that for more complex multi-step workflows, text descriptions often miss edge cases and decision point
16.
▲
by
YaraDori
8mo ago
Really interesting approach to the coordination problem. File locking + knowledge flywheel is a clean pattern. I've been thinking about the adjacent problem: even once you solve coordination, each agent still needs to know HOW to do th
17.
▲
by
YaraDori
8mo ago
This is a great approach — giving agents their own compute environment is a huge unlock. One challenge I keep running into on the other side of this: even with a full VM and browser available, the bottleneck is often teaching the agent what
18.
▲
Show HN: SkillForge – Turn Screen Recordings into Agent-Ready Skills
(skillforge.expert)
1 points
by
YaraDori
8mo ago
|
2 comments