theredbeard

4 ms·

1.
by theredbeard 6mo ago
We haven’t been inching closer to users writing a half-decent ticket in decades though.
2.
by theredbeard 6mo ago
This is because for some reason all agentic systems think that slapping cron on it is enough, but that completely ignores decades of knowledge about prospective memory. Take a look at https://theredbeard.io/blog/the-mis
3.
by theredbeard 6mo ago
What in the ChatGPT is this?
4.
by theredbeard 6mo ago
I’m getting pretty decent at spotting LLM text. This doesn’t contain the obvious tells at least.
5.
by theredbeard 7mo ago
It’s a self fulfilling prophecy. They’re extremely expensive so they must be good so they must be worth it. And because at that level measurement is extremely subjective it’s mainly about the vibes. Like everything it’s just marketing.
6.
by theredbeard 7mo ago
I’m sorry but no attempt was made here. It contains all the red flags in the first few paragraphs.
7.
by theredbeard 7mo ago
A vibe? It’s completely obvious AI slop with no attempt to make it legible. They didn’t even prompt out the emdashes. For such a cool finding this is extremely disappointing.
8.
by theredbeard 7mo ago
It's a fair question. I've had problems with Gemini 3 due to rate limiting, and I've been working on this for a while now. I'm planning Gemini 3 for a follow up.
9.
I traced 3,177 API calls to see what 4 AI coding tools put in the context window(theredbeard.io)
36 points by theredbeard 7mo ago|4 comments
10.
Show HN: Context Lens: View your CLI's agent context in realtime(github.com)
1 points by theredbeard 7mo ago|0 comments
11.
by theredbeard 7mo ago
It’s not groundbreaking in a technological sense. The codebase is actually a bit of a monstrosity. But it removed guardrails that were artificially put on these LLMs which suddenly gave it an entire new dimension and the timing was right.
12.
by theredbeard 8mo ago
I built this because I was curious what Claude sends to the API, how subagents get work delegated and what contexts look like. Interesting to see how small part of the context the user interaction really is typically.
13.
Show HN: Context Lens: Devtools for your agent context(github.com)
3 points by theredbeard 8mo ago|2 comments
14.
Are we becoming QA for the machine?(medium.com)
1 points by theredbeard 8mo ago|0 comments
15.
by theredbeard 8mo ago
I built this because I was curious what Claude sends to the API, how subagents get work delegated and how contexts look like. Interesting to see how small part of the context the user interaction really is typically.
16.
Show HN: Context Lens – See what's inside your AI agent's context window(github.com)
2 points by theredbeard 8mo ago|1 comments
17.
by theredbeard 8mo ago
No worries, the italics did heavy lifting.
18.
by theredbeard 8mo ago
Gitlab.com is the obvious rec.
19.
by theredbeard 8mo ago
What could go wrong?!
20.
by theredbeard 8mo ago
Skipping the investigation phase to jump straight to solutions has killed projects for decades. Requirements docs nobody reads, analysis nobody does, straight to coding because that feels like progress. AI makes this pattern incredibly attr
21.
by theredbeard 8mo ago
OSS was already brutal for new contributors before AI. You'd spend hours on a good-faith PR and get ignored for months, or get torn apart in review because you didn't know the unwritten conventions. The signal-to-noise ratio sucke