redhale

9 ms·

91.
by redhale 1y ago
This is interesting, and I think there are use cases where this feature may make sense. But this is not the feature they should or could have built, at least for Claude Code. CC already had a feature very similar to this -- subagents (or ag
92.
by redhale 1y ago
Do you have a study to back that up? /s I agree. I think there are too many resources, examples, and live streams out there for someone to credibly claim at this point that these tools have no value and are all hype. I think the nuance
93.
by redhale 1y ago
I think the idea would be that you could re-compact with a different focus. When you compact, you can give Claude instructions on what is important to retain and what can be discarded. If you later discover that actually you wanted somethin
94.
by redhale 1y ago
Yes, but the difference is that an LLM produces the result instantly, whereas a human might take hours or days. So if you can get the spec right, and the LLM+agent harness is good enough, you can move much, much faster. It's not always
95.
by redhale 1y ago
Is it really a mystery? Perhaps stereotypes and political majorities don't accurately describe anywhere near 100% of the people they're applied to? By a similar standard, the US as a whole is anti-intellectual. And yet...
96.
by redhale 1y ago
In my opinion, subagents (or more generally, "agents as tools" as a pattern) are an order-of-magnitude level feature. Soon every CLI agent will have them as a first-class feature (you can get them via custom scripting right now wi
97.
by redhale 1y ago
My guess is: absolutely not, at least not for more than a few minutes. Subagents chew through tokens at a very high rate, and this system makes heavy use of subagents.
98.
by redhale 1y ago
I agree, I think traceability will be extremely important in evolving and improving a system like this. Since scripting is involved in searching for and managing skills, I feel like there is probably a way to achieve some kind of use tracin
99.
by redhale 1y ago
I don't necessarily think the conclusions are wrong, but this relies entirely on self-reported survey results to measure productivity gains. That's too easy to poke holes in, and I think studies like this are unlikely to convince
100.
by redhale 1y ago
Subagents are a critical feature that GH Copilot still lacks. They allow your main agent to use another agent as a tool, meaning the main agent's context doesn't get nearly as polluted. Good read on the benefits of this pattern:
101.
by redhale 1y ago
Everything is just context, of course. Every time I see a blog post on "the nine types of agentic memory" or some such I have a similar reaction. I would say that systems like this are about getting the agent to correctly choose t
102.
by redhale 1y ago
I love this type of tool, and I like that there are several people and companies taking a stab at it. Either of you or any other author of a similar tool: please add support for OpenCode [0]! [0] https://opencode.ai/
103.
by redhale 1y ago
Thanks for sharing! The problem with every single tool in the category that I've come across (e.g. Conductor, Sculptor) is that they assume a single repository. Very rarely in my career working on enterprise software have I been in a s
104.
by redhale 1y ago
Totally agree. The CLI agents (or whatever) lower the barrier of entry to building a custom agent all the way down to basically just writing markdown. Excellent article in this vein: https://jxnl.co/writing/2025/09
105.
by redhale 1y ago
I think seatbelts or airbags are better analogies. And both seatbelts and airbags are mandated in the US. Of course saving lives isn't the ONLY metric (perhaps all personal vehicles would be banned if that were the case), but we make r
106.
by redhale 1y ago
Agree. This may seem nice and easy for simple demo scenarios, but things always get more complicated ("I just need an if statement here and a switch statement there") and imo code always wins in the end. Perhaps for truly pure dat
107.
by redhale 1y ago
This looks awesome! I really hope there is planned support for other coding agents too, in particular OpenCode which seems to have relatively close feature parity coupled with wide model compatibility and open source.
108.
by redhale 1y ago
Even on Windows, I use Rider. It's just better imo, and it's not close. But then again I have not used VS deeply in a couple years.
109.
by redhale 1y ago
> The only reason it works for many common programming workflows is because they are common. The minute you try to do new things, you need to be as verbose as the underlying language. You can stop here. This is enough to change the game.
110.
by redhale 1y ago
This! Subagents open all the new metaphorical tabs to get to some answer, then close those tabs so the main agent can proceed with the main task. Excellent article on this pattern: https://jxnl.co/writing/2025/08&#
111.
by redhale 1y ago
I don't know why people are still eager to spike the ball so quickly after every wave of progress. Things are still moving SO FAST. ---------------- Ok it can complete a line of python, it will never write a full, correct function. Ok
112.
by redhale 1y ago
I recently saw SemTools [0], but have not tried it out yet myself. [0] https://github.com/run-llama/semtools
113.
by redhale 1y ago
Honest feedback: I am usually not someone obsessed with privacy. I use Google services for a lot of things, I accept most cookies, I don't use local models, etc. But when I read "captures screenshots every 5 seconds while you work
114.
by redhale 1y ago
I have found that adding things to a todo (Todist) list called "want-to-buy" gives me the little endorphin boost / anxiety relief without spending any money. I periodically go through and delete stuff from this list after tim
115.
by redhale 1y ago
> It’s always easier to straighten a sapling than a grown tree. Ok, but in this case you can just throw away the tree and a new one will grow immediately for you to review anew. Rinse and repeat. I'm not saying the author's pro
116.
by redhale 1y ago
100,% this. PAYG bill anxiety is such a barrier that I will not use it AT ALL if I'm paying the bill. And even when my company is paying, I feel a need to be too careful as to negate much of the benefit. Raise the prices of the all-inc
117.
by redhale 1y ago
In my case I'm creating the agent in a custom way (using the BAML framework), so I don't need the agent part of your solution. FastMCP works ok but has some issues, so if you have a better client than can be used as a replacement,
118.
by redhale 1y ago
I'm confused -- I didn't see a single example in you readme or docs that uses the client WITHOUT your agent. Trying to understand if this is an alternative to FastMCP. I don't want your agent.
119.
by redhale 1y ago
It's clearly satire.
120.
by redhale 1y ago
This is just an ad for Kilo Code with a very dumb take. Claude Max, to my knowledge, was never marketed as "unlimited". Claude Max gives you WAY more tokens then $100/$200 would buy. When you get rate limited, you have the op

More ›