aluzzardi

7 ms·

1.
by aluzzardi 5mo ago
Thank you, appreciate it! Regarding scoping: In our case, the agent loop runs in the same way as our API server does (as in, it’s a multi tenant service running in a container somewhere). And we solve scoping in the same way. To put it in o
2.
by aluzzardi 5mo ago
Author here. I should have made it more clear that the article is about agent / harness building (not about running third party agents). > I barely trust the harness more than the LLM Since we built it, I trust it just as much as I
3.
by aluzzardi 5mo ago
Author here. I’m worried about the same (models tuned for specific harnesses). We actually work around that by respecting the “contract”. For instance, our harness’ Bash signature is exactly the same as Claude’s. We do our sandboxing stuff
4.
by aluzzardi 5mo ago
Author here. This is an interesting and novel field, so I’m not pretending I know the answers, but this is what worked for us :) At the end of the day, and oversimplifying things: why would I want to spawn a for loop that calls an API (LLM)
5.
by aluzzardi 5mo ago
Author here. I think the confusion is that “agent” is used for two very different things: - building an agent - an “agent” product/runtime (Claude Code, etc) In the first case, the model never executes anything. It just outputs somethi
6.
by aluzzardi 5mo ago
Author here. In my opinion, the main driver here is how fast models have evolved in the past 12 months. It makes the architecture of everything around them obsolete, very fast. We went from using models as a building block, wrapping them in
7.
by aluzzardi 5mo ago
Author here. Because of parallelism and non determinism. This problem is quite common and not limited to memories. For instance, Claude Code will block write attempts and steer the agent to perform a read first (because the file might have
8.
by aluzzardi 5mo ago
Author here. My definition is: you take an agent, remove the model and you’re left with the harness. Tools, memories, sandboxing, steering, etc
9.
by aluzzardi 5mo ago
Author here. Depending on how it’s designed, the harness itself doesn’t need any sandboxing. At the end of the day, it’s a “simple” loop that calls an external API (LLM) and receives requests to execute stuff on its behalf. It’s not the age
10.
We Built Our AI Agent(mendral.com)
2 points by aluzzardi 6mo ago|0 comments
11.
Our Agent's Most Important Job Is Deciding Not to Think(mendral.com)
4 points by aluzzardi 7mo ago|0 comments
12.
by aluzzardi 7mo ago
It started with Sonnet 4.0 as a single agent and now it’s a mix of Opus 4.6 and Haiku 4.5 agents. Opus plans the investigation and orchestrates the searches. Haiku is the one actually querying ClickHouse and returning relevant bits
13.
by aluzzardi 7mo ago
> it's not magic and you need to make the job of the agent easier by giving it good instructions, tools, and environments. This. We had much better success by letting the agent pull context rather trying to push what we thought was
14.
by aluzzardi 7mo ago
There are 2 layers of compression: - ZSTD (actual data compression) - De-duplication (i.e. what you're saying) Although AFAIK it's not "just point to it" but rather storing sorted data and being able to say "the nex
15.
by aluzzardi 7mo ago
Mendral co-founder and post author here. I agree with your statement and explained in a few other comments how we're doing this. tldr: - Something happens that needs investigating - Main (Opus) agent makes focused plan and spawns sub a
16.
by aluzzardi 7mo ago
From our experience running this, we're seeing patterns like these: - Opus agent wakes up when we detect an incident (e.g. CI broke on main) - It looks at the big picture (e.g. which job broke) and makes a plan to investigate - It disp
17.
by aluzzardi 7mo ago
> My experience with LLM generated SQL in OLTP and OLAP platforms has been a mixed bag Models are evolving fast . If your experience is older than a few months, I encourage you to try again. I mean this with the best intentions: it&#x27
18.
by aluzzardi 7mo ago
We've actually started to gather metrics this week to write that exact post :) Coming soon!
19.
by aluzzardi 7mo ago
Mendral co-founder here and author of the post. This is an interesting approach. I definitely agree with the problem statement: if the LLM has to filter by error/fatal because of context window constraints, it will miss crucial informa
20.
by aluzzardi 7mo ago
Post author here. Yes, it works really well. 1) The latest models are radically better at this. We noticed a massive improvement in quality starting with Sonnet 4.5 2) The context issue is real. We solve this by using sub agents that read t
21.
LLMs Are Good at SQL. We Gave Ours Terabytes of CI Logs(mendral.com)
19 points by aluzzardi 7mo ago|0 comments
22.
Show HN: Container Use for Agents(github.com)
82 points by aluzzardi 1y ago|17 comments