Skip to content
View ondraulehla's full-sized avatar
3 followers · 46 following

Achievements

Achievement: Pair Extraordinairex3Achievement: YOLOAchievement: Pull Shark

Achievements

Achievement: Pair Extraordinairex3Achievement: YOLOAchievement: Pull Shark

Block or report ondraulehla

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.

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
ondraulehla/README.md

Ondřej Úlehla

Fullstack developer from Prague. Right now most of my work is automating business processes with AI agents, above all building the harness they run on: skills, rules, subagents and scripts, MCP servers that give them tools, and RAG indexes I then wire the agents into. Plenty of straight fullstack comes with it, along with sitting down with clients to work out which problems are worth automating at all. In my free time I build open-source tooling for AI agents.

🌐 ulehla.dev · 💼 LinkedIn

Projects

  • agent-flight-recorder: run a task N times in isolated E2B sandboxes, record full traces, measure pass rate, flakiness and cost, and diff trajectories
  • readme-ci: run the code blocks in your README and fail CI when your quickstart breaks
  • agent-audit: security auditor for AI agent configurations, covering MCP servers, Claude Code settings, skills and hooks
  • agent-lens: drop an agent session transcript and get a visual debrief with tool-call timeline, tokens, cost and failures
  • mcp-sandbox: run any MCP server inside an isolated E2B cloud sandbox with one command, no code changes

Pinned Loading

  1. agent-flight-recorder agent-flight-recorder Public

    Flight recorder for AI agents: run tasks N times in isolated E2B sandboxes, record full traces, measure pass rate, flakiness and cost, and diff trajectories

    Python

  2. agent-audit agent-audit Public

    Security auditor for AI agent configurations — scans MCP servers, Claude Code settings, agents, skills and hooks for risky permissions, inline secrets, unpinned supply chains and prompt-injection s…

    TypeScript

  3. agent-lens agent-lens Public

    See what your agent actually did — drop an AI agent session transcript (JSONL) and get a visual debrief: tool-call timeline, tokens, cost, failures. 100% client-side.

    TypeScript

  4. mcp-sandbox mcp-sandbox Public

    Run any MCP server inside an isolated E2B cloud sandbox instead of on your machine — one command, no code changes. Transparent stdio bridge, env allowlisting, JSONL tracing.

    TypeScript

  5. url-shortener-microservices url-shortener-microservices Public

    Production-style URL shortener as microservices — cache-first redirects, Redis Streams analytics, Prometheus/Grafana observability, Docker Compose + k8s reference manifests. 56 tests, no infra need…

    JavaScript

  6. readme-ci readme-ci Public

    Run the code blocks in your README and fail CI when your quickstart breaks

    TypeScript