Skip to content
View jLAM-ERR's full-sized avatar
7 followers · 1 following

Achievements

Achievement: Pull Sharkx2Achievement: Pair Extraordinairex2

Achievements

Achievement: Pull Sharkx2Achievement: Pair Extraordinairex2

Organizations

@CetaLab

Block or report jLAM-ERR

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
jLAM-ERR/README.md

Hi, I'm Artem

Engineering manager with a backend background. I run a Java department of 18 engineers at a FinTech company and still build things myself — mostly around how AI coding agents fit into real engineering teams.

What I care about

  • Letting developers use AI coding agents inside regulated environments — without leaking what shouldn't leave the building
  • Making agent work compound: knowledge that survives session resets instead of dying in chat scrollback
  • Reactive, high-load JVM systems (Spring WebFlux, Project Reactor, Kafka)

Projects

Project What it does
corp-llm-gateway Corporate LLM gateway that sanitises traffic between AI coding agents (Claude Code, Codex, Cursor) and Anthropic / OpenAI before it leaves the corporate boundary. Local detection cascade (~6 ms p50): Russian entity checksums, bilingual NER, AML/CFT gazetteers, secret patterns. Two-layer DLP, Keycloak auth, audit into Langfuse / S3 / SIEM, Helm chart.
autodidact A knowledge base that makes AI coding agents keep what they learn. Three context tiers keep always-loaded instructions tiny, a capture → retro → distill → promote loop turns review-gate decisions into reusable knowledge, and a lint enforces size budgets and a PII scan. Portable across Claude Code, Codex, OpenCode and Kilo Code. Paired with autodidact-plugins — workflow and learning-loop plugins for the Claude Code marketplace.
corvex Cross-platform CLI for Xray and AmneziaWG tunnels: VLESS / VMess / Trojan / Shadowsocks, subscription auto-discovery with health checks, one-command system proxy on macOS, Linux and Windows.

Pinned Loading

  1. corp-llm-gateway corp-llm-gateway Public

    Corporate LLM gateway. Sanitizes traffic between developer Claude Code instances and Anthropic/OpenAI before it leaves the corp boundary.

    Python 4 2

  2. autodidact autodidact Public

    autodidact — what a coding agent teaches itself: a knowledge base that makes AI coding agents keep what they learn (reviewed knowledge units, project skeleton, methodology, versioned contract)

    Python

  3. corvex corvex Public

    Cross-platform CLI for Xray & AmneziaWG tunnels: VLESS, VMess, Trojan, Shadowsocks, subscription auto-discovery with health checks, and one-command system proxy setup on macOS, Linux, and Windows.

    Rust 1