Currently we're at 1.8 CVEs per day since OpenClaw launched!

Superpowers has several skills. Its core workflow is:

- Brainstorm with you to design a spec

- Use subagents to review its own spec, then get your approval

- Based on the spec, write a plan, use subagents to review before final approval

- Use subagents to implement (using TDD)

I think that the brainstorming skill [1] is great. It helps flesh out a rough early idea. I also like that it uses subagents to adversarially review its own spec/plan; that has caught several things I would've missed. I do not like the separation of spec/plan; IMO the models are good enough to get straight to coding once the spec is written. The plan often ends up being code blocks in a Markdown doc.

[1]: https://github.com/obra/superpowers/tree/main/skills/brainst...

Comments URL: https://news.ycombinator.com/item?id=47428345

Points: 2

# Comments: 0

The US has nearly always been successful in terms of conventional firepower and individual operations. E.g. in 2003 the US overthrew Saddam's government in a matter of weeks. The US won most battles in Vietnam. That doesn't change the fact that the strategic outcomes and long-term track record are poor. Trying to draw a link to AI or the current state of the US military feels flimsy.

Anyway, the recurring Big Question throughout the Culture series is "how should a highly progressive, developed, and egalitarian society act when it meets others who are not?". The US is sliding further and further from that ideal, and you can argue whether it was ever close.

I've been using Claude Code etc. for nearly all my work lately, and I wanted to see how many of my commits it was actually co-authoring. So I made this little tool to visualize my usage.

You log in with Github (read-only), it scans your commits from the last year, and visualizes how many came from Claude, Cursor, or any tool that adds Co-Authored-By trailer to the commit message.

Caveat: not all tools add this trailer, so this doesn't include e.g. Codex. Still, hope you find it interesting!

Comments URL: https://news.ycombinator.com/item?id=47247102

Points: 5

# Comments: 2

Nanobot's was not great (cron + a HEARTBEAT.md meant two ways to do things, which would confuse the AI). But because the implementation is so simple, I could improve it in a few minutes in my own fork!

Not exactly symbolic AI, but pretty cool nonetheless.

- OpenClaw: the big one, but extremely messy codebase and deployment

- NanoClaw: simple, main selling point is that agents spawn their own containers. Personally I don't see why that's preferable to just running the whole thing in a container for single-user purposes

- IronClaw: focused on security (tools run in a WASM sandbox, some defenses against prompt injection but idk if they're any good)

- PicoClaw: targets low-end machines/Raspberry Pis

- ZeroClaw: Claw But In Rust

- NanoBot: ~4k lines of Python, easy to understand and modify. This is the one I landed on and have been using Claude to tweak as needed for myself

Comments URL: https://news.ycombinator.com/item?id=45633313

Points: 1

# Comments: 0

You could try to get your agent to test its own skills. From https://blog.fsck.com/2025/10/09/superpowers:

> As Claude and I build new skills, one of the things I ask it to do is to "test" the skills on a set of subagents to ensure that the skills were comprehensible, complete, and that the subagents would comply with them. (Claude now thinks of this as TDD for skills and uses its RED/GREEN TDD skill as part of the skill creation skill.)

> The first time we played this game, Claude told me that the subagents had gotten a perfect score. After a bit of prodding, I discovered that Claude was quizzing the subagents like they were on a gameshow. This was less than useful. I asked to switch to realistic scenarios that put pressure on the agents, to better simulate what they might actually do.

[^1]: https://www.youtube.com/watch?v=3sVO_j4czYs

Comments URL: https://news.ycombinator.com/item?id=45412573

Points: 3

# Comments: 1