anatolii1309

The root issue is not the button, it is that the allowlist stores the full command string, including whatever was in it. We ran into this ourselves: approve `curl -H "Authorization: Bearer sk-..."` once and that token sits in settings.local.json until someone happens to reread the file.

Two rules that stuck for us: secrets go into commands only via env vars ($TOKEN, never the value), and .claude/ lives in the machine-wide global gitignore, not in the repo. With those, "Don't ask again" is fine and a commit has nothing to leak.

Location: Warsaw, Poland (CET) Remote: yes Willing to relocate: no Technologies: Python (FastAPI, Django), TypeScript (React/Next.js, Node.js), PostgreSQL, Docker, AWS/GCP/Azure, LLM/agent systems (RAG, LangGraph, MCP) Résumé/CV: https://linkedin.com/in/anatoli-navahrodski-55854949 | GitHub: https://github.com/Anatolii1309 Email: anatolii1309@gmail.com

Technical co-founder / partner, not a job seeker.

16+ years building software, two engineering practices built from zero (15+ engineers), running my own software company since 2014, production AI systems since 2023 (RAG, multi-agent, LangGraph).

What I bring: turning an idea into a working product and a team — architecture, hiring, process, security, compliance (GDPR, EU AI Act).

Proof of work: SealedRun — open-source tamper-evident audit trail for AI agents (Ed25519 + ML-DSA-65, Apache-2.0, on PyPI/npm): a signed, offline-verifiable record of every agent step. https://github.com/sealedrun/sealedrun / https://sealedrun.com

Looking for: an equity partnership in an early-stage B2B / AI product that needs a technical co-founder; equity + cash. Strongest fit: regulated markets (fintech, healthtech, legaltech, AI governance).