alexsmirnov

5 ms·

1.
by alexsmirnov 20d ago
I do have SUV ( Pathfinder ) for camping, skiing, kayaking, and other long trips. And 2 seats electric Smart for local commute - groceries, school drop off/pickup, gym visits etc. This is 95% of my family car use. The 90 miles charge i
2.
by alexsmirnov 26d ago
I did create special repo that has all skills/agents/scripts and MCP configuration. All agents ( currently 3 in use: claude code, opencode, and pi ) packed in docker container, with artifacts required by project technologies, and
3.
by alexsmirnov 1mo ago
In my feeling, it is decades old. With LLMs and agentic coding, it's deja vu of some 200x working with with inexperienced offshore teams. The same misunderstanding problems, the same corners cut, the same attempts to present bullshit a
4.
by alexsmirnov 2mo ago
"The single worst issue: the system prompt never explicitly tells it to ignore auto-injected files like CLAUDE.md or *.rule.md." - Yes, it is! Right after the content of CLAUDE.md, Claude code inserts: <system-reminder> IMP
5.
by alexsmirnov 2mo ago
I did run both kimi-k3 and glm-5.2 with Capital One vulnhunt [1]. No rejections, they did find the same problems on my project that I used for testing. gpt-5.6, gemini pro, and opus all rejected to follow. gpt even declined to edit skill fi
6.
by alexsmirnov 3mo ago
Shit happens. I had my iMac and Time machine hard drives died 2 days apart...
7.
by alexsmirnov 3mo ago
The problem of duplicated code described in the article, goes in a different way in reality: AI does not update 4 places in the same way, but implement them a slightly different. I found diverged business logic all the time. For example, fi
8.
by alexsmirnov 3mo ago
> This happens but far less often than it used to, and the case for full autonomous agents is getting stronger, not weaker. This is that I do not see. My journey, just couple weeks ago, Claude Code + Opus 4.8. The task was not too compli
9.
by alexsmirnov 3mo ago
If I pay for vending machine by corporate card on a business trip, it looks more like B2B
10.
by alexsmirnov 3mo ago
From the first hand: I was in the air base service regiment, Ovruch ( 30 km from reactor ), 1 year in service at the time of disaster. The average numbers have a little sense. By our measurements, it was't an "even" distribut
11.
by alexsmirnov 4mo ago
I more like my high school math teacher explanation: "Put your one hand on dry ice, another in the boiling water. On average, you feel warm and cozy"
12.
by alexsmirnov 4mo ago
We usually squash feature branches before merge. To squash before rebase, I use git reset --soft $(git merge-base develop HEAD) && git commit && git rebase develop - you have to resolve final conflicts only
13.
by alexsmirnov 6mo ago
Much longer than that, and was available way before an internet. I graduated STEM high school in St. Petersburg in 1981, and I had several classmates who were big funs of chemistry. That they were able to create from textbooks, school lab i
14.
by alexsmirnov 6mo ago
Exactly. I did create my own MCP with custom agents that combine several tools into a single one. For example, all WebSearch, WebFetch, Context7 exposed as a single "web research" tool, backed by the cheapest model that passes eva
15.
by alexsmirnov 7mo ago
Almost instantly, compared to my experience working for a big health care provider... I waited 6 moths for IT department to allow me install development tools on work laptop. And while security rules created enormous roadblocks for work, wh
16.
by alexsmirnov 7mo ago
> you start off checking every diff like a hawk, expecting it to break things, but honestly, soon you see it's not necessary most of the time. I see it's necessary ALL the time. The AI generated code can be used as scaffolding,
17.
by alexsmirnov 7mo ago
The calculation misses subagents tokens, that can be a significant differences. Better to parse session jsonl files ~/.claude/projects/<project>/<session id>.jsonl and ~/.claude/projects/<pr
18.
by alexsmirnov 7mo ago
This mostly matches my own estimates for pr-review command that I use. But it's pretty sophisticated: 6 specialized agents, best practices skills, CVE database, bunch of scripts. To reduce cost, most of agents use cheap open source mod
19.
by alexsmirnov 7mo ago
Actually, they extremely bad at that. All training data contains cod + tests, even if tests where created first. So far, all models that I tried failed to implement tests for interfaces, without access to actual code.
20.
by alexsmirnov 7mo ago
Considering how little data needed to poison llm https://www.anthropic.com/research/small-samples-poison , this is a way to replace SEO by llm product placement: 1. create several hundreds github repos with projects th
21.
by alexsmirnov 7mo ago
I do in similar way, connect claude code to litellm router that dispatches model requests to different providers: bedrock, openai, gemini, openrouter and ollama for opensource models. I have special slash command and script that collect inf
22.
by alexsmirnov 8mo ago
For me, AI is the best for code research and review Since some team members started using AI without care, I did create bunch of agents/skills/commands and custom scripts for claude code. For each PR, it collects changes by git lo
23.
by alexsmirnov 8mo ago
I did create and actively use a similar tool, but with different purpose: configure AI tools for each team member to use the same code style and architecture guides across projects. It includes: - build docker images for claude code and ope
24.
by alexsmirnov 8mo ago
I do use Obsidian on pair with Claude code and git. I organize notes by tags, folders, and links from tree of "map of content" notes. Those documented as rules for AI. All notes came to "Inbox" folder, and from time to
25.
by alexsmirnov 9mo ago
This was implemented far ago, at least by huggingface "smolagents". https://huggingface.co/docs/smolagents/index . I did use them, with evaluations. For the most cases, modern models tool call outperform
26.
by alexsmirnov 9mo ago
This is exact the impression that I got. Every question or task given to LLM returns pretty reasonable, but flawed result. For the coding, those are hard to spot but dangerous mistakes. They all look good and perfectly reasonable, but just
27.
by alexsmirnov 9mo ago
This is not about understanding the message, but switching user mental activity. I go myself in the similar situations many times. One example: I tried to pay my bills in online bank application, but got into error. After several attempts,
28.
by alexsmirnov 10mo ago
A lot of discussions around vibe AI coding flaws: awful architecture, performance problems, security holes, lack of maintainability, bugs, and low code quality. All correct, but none of those is matter if: - you create small utility that co
29.
by alexsmirnov 10mo ago
I've read book written by captain Kocebu, that was on duty to protect Russian holdings in Alaska. They visited San Francisco in 1805 and 1815, and several chapters described life of native people in the mission. He described harsh cond
30.
by alexsmirnov 11mo ago
I saw it more and more recent times, but worse is coming: With my teenage daughter ( and her friends ), I see that they even do not bother with screenshots, and take pictures by phone...

More ›