HackerTrans
TopNewTrendsCommentsPastAskShowJobs

sshh12

no profile record

Submissions

Designing Software for Software Factories

blog.sshh.io
3 points·by sshh12·28 hari yang lalu·0 comments

Waypoint-1.5

over.world
4 points·by sshh12·bulan lalu·0 comments

WikiProject Editor Retention

en.wikipedia.org
2 points·by sshh12·2 bulan yang lalu·1 comments

[untitled]

18 points·by sshh12·2 bulan yang lalu·0 comments

AI drives companies to reorganize around loops, not functions

blog.sshh.io
4 points·by sshh12·3 bulan yang lalu·0 comments

Show HN: Brw – Browser automation for Claude Code agent teams

github.com
1 points·by sshh12·4 bulan yang lalu·0 comments

Plan.md

github.com
1 points·by sshh12·4 bulan yang lalu·0 comments

Revealed Preferences

writing.nikunjk.com
2 points·by sshh12·4 bulan yang lalu·1 comments

Our Design Docs Write Themselves

abnormalbuilders.substack.com
2 points·by sshh12·4 bulan yang lalu·0 comments

Reverse-engineering how Claude's Chrome extension controls the browser

gist.github.com
3 points·by sshh12·4 bulan yang lalu·3 comments

Taste Is Not a Moat

blog.sshh.io
3 points·by sshh12·5 bulan yang lalu·0 comments

Token Anxiety

writing.nikunjk.com
9 points·by sshh12·5 bulan yang lalu·1 comments

Move Faster

blog.sshh.io
2 points·by sshh12·6 bulan yang lalu·0 comments

Show HN: Modal Agents SDK

github.com
4 points·by sshh12·6 bulan yang lalu·0 comments

RCS for Business

developers.google.com
56 points·by sshh12·6 bulan yang lalu·65 comments

All agents will become coding agents

davistreybig.substack.com
1 points·by sshh12·6 bulan yang lalu·0 comments

Agents Should Write Code, Not JSON

blog.sshh.io
3 points·by sshh12·6 bulan yang lalu·4 comments

Most Code Is Just Cache

blog.sshh.io
3 points·by sshh12·6 bulan yang lalu·0 comments

Time Ablation Experiments on tau2-bench

github.com
3 points·by sshh12·6 bulan yang lalu·0 comments

Show HN: Train Claude Skills on Your PR History

github.com
2 points·by sshh12·6 bulan yang lalu·0 comments

comments

sshh12
·2 bulan yang lalu·discuss
Had codex whip up a ./claude-via-codex.sh for resuming claude cli sessions in the meantime:

https://gist.github.com/sshh12/f2c402db02eb00982100f645519bd...
sshh12
·2 bulan yang lalu·discuss
Oof well I guess I'm personally a fan of buzz feed listicles. Thats been my style pre-AI.
sshh12
·2 bulan yang lalu·discuss
Any issues with the content itself?

It shouldn't be that surprising that a post advocating for using more AI is written using AI.
sshh12
·4 bulan yang lalu·discuss
slack: https://gist.github.com/sshh12/4cca8d6698be3c80e9232b68586b7...
sshh12
·4 bulan yang lalu·discuss
netflix: https://gist.github.com/sshh12/dda3a89514f850c459380b18b1f7e...
sshh12
·4 bulan yang lalu·discuss
This was literally just a single opus prompt but thought it was pretty interesting how Anthropic is designing browser-based harnesses under the hood.
sshh12
·6 bulan yang lalu·discuss
TIL they have a rich multimedia API for RCS.
sshh12
·6 bulan yang lalu·discuss
ack thanks -- didn't realize this was against the guidelines
sshh12
·6 bulan yang lalu·discuss
Youre right and there are some assumptions being made here around the agent having enough context to work on a task without interrupts (e.g. team review, asking questions, etc).

Typically human equivalent time is based on a single person given all the potential information they need up front (which is not today how a lot of work is done).
sshh12
·7 bulan yang lalu·discuss
For folks interested in some of the nuances of this benchmark, I just posted this deep dive:

https://blog.sshh.io/p/understanding-ai-benchmarks
sshh12
·8 bulan yang lalu·discuss
IMO MCP isn't totally dead, but its role has shrunk. Quoting from my post [1]:

"Instead of a bloated API, an MCP should be a simple, secure gateway... MCP’s job isn’t to abstract reality for the agent; its job is to manage the auth, networking, and security boundaries and then get out of the way."

You still need some standard to hook up data to agents esp when the agents are not running on your local dev machine. I don't think e.g. REST/etc are nearly specific enough to do this without a more constrained standard for requests.

[1] https://blog.sshh.io/p/how-i-use-every-claude-code-feature
sshh12
·8 bulan yang lalu·discuss
All markdown files, yeah
sshh12
·8 bulan yang lalu·discuss
We have a linter that checks for this to help mitigate
sshh12
·8 bulan yang lalu·discuss
Yeah I'm fairy pessimistic about how much folks will read
sshh12
·8 bulan yang lalu·discuss
Hooks can also be useful for this. If it's using the wrong APIs then can hint on write or block on commit with some lint function that checks for this.
sshh12
·8 bulan yang lalu·discuss
Yeah I started with Cursor, went hybrid, and then in the last month or so I've totally swapped over.

Part of it is the snappy more minimal UX but also just pure efficacy seems consistently better. Claude does its best work in CC. I'm sure the same is true of Codex.
sshh12
·8 bulan yang lalu·discuss
Yeah if that's possible or you are willing to build it, that's the right solution. Today pretty much all of my integrations are pure CLIs like that rather than MCPs.

You can do anything you want via a CLI but MCP still exists as a standard that folks and platforms might want to adopt as a common interface.
sshh12
·8 bulan yang lalu·discuss
Often code in SaaS companies like ours is indeed how we solve customer problems. It's not so much the amount of code but the rate (code per time) we can effectively use to solve problems/build solutions. AI, when tuned correctly, lets us do this faster than ever possible before.
sshh12
·8 bulan yang lalu·discuss
Makes sense. I work for a growth stage startup and most of these apply to our internal mono repo so hard to share specifics. We use this for both new and legacy code each with their own unique AI coding challenges.

If theres enough interest, I might replicate some examples in an open source project.
sshh12
·8 bulan yang lalu·discuss
I've found planning to be key here for scaling to arbitrary complex changes.

It's much easier to review larger changes when you've aligned on a Claude generated plan up front.