Aherontas

How I Built a Signed Audit Trail for an LLM System (And Broke It With One SQL Update)

Logs, traces, governance, compliance, audit trails... those are some of the words that come up every time we discuss putting LLM features and AI agents into production.

Most systems already have logs. The real question is different:

Can you prove that the logs you are showing today are the same records the system wrote six months ago?

Part 1/7 of my PyCon Greece 2026 talk, From Prompt to Proof: a Presidio guardrail masked the ΑΦΜ(tax number in Greece) label but leaked the actual tax identifier to the LLM while returning HTTP 200.

Unfortunately we will see this kind of cases more and more with AI rise. I don't believe it is the only app that could do relevant labeled searching in faces etc.

A same to not have them anymore :P

Curious how you handle feed evolution over time. When an RSS source changes structure (fields added/removed, summaries truncated, etc.), do you normalize to a fixed schema or store the raw payload alongside a best-effort normalized version? Longitudinal datasets tend to get tricky there.

I am doing exactly the same, also writing down some raw ideas I have in it to not forget them.

Just released it if you are interested: https://www.petrostechchronicles.com/blog/PyCon_Greece_2025_...

If you are interested, I just released the slides in my blog: https://www.petrostechchronicles.com/blog/PyCon_Greece_2025_...

That's correct, the slides are mostly showing how MCP - A2A work and basic architectures between them and Agents in general. I will upload probably today the presentation slides in my blog.

Hello exactly what you are saying is what this repo does. I will upload today probably the whole PyCon presentation of it in my blog if you are interested. You will find it at: https://www.petrostechchronicles.com/

It just builds a ChatGPT like UI that has Agents to serve you either latest news on a specific tech topic (via fetching them via MCP Servers for: Brave Browser, Hackernews and Github for trending repos or repos fetched in the response). It also supports giving you back general knowledge answers. The whole purpose of the repo is to be a general example of how Agents speak with Agents and how to pair them with MCP servers and why we need MCP Servers(for example better run all calculations not in LLM because they hallucinate, but via code and then give the answer back to the LLM). I hope it clarifies what it does now.

Haven't tried balm, I will give it a shot, I am really curious to see all the features it supports.

Thanks a lot for taking the time to go through the code and leave such a detailed comment, I really appreciate the thoughtful feedback either bad or good, it always helps.

Just to clarify: this repo is meant as a quick demo/learning resource after experimenting with Pydanticai(for some limited time), FastAPI, and agents. It’s not something I’d structure or ship in a production environment or at a real job in any case.

That said, your points are spot on:

- Consistency in return types (Pydantic schemas over mixed dicts/JSONResponse, which was done either because I glued code from other projects I had either from generated so if it is used from anyone in real case it needs refactoring).

Structuring data exchange between agents with typed models instead of raw dicts. (totally correct too)

Avoiding redundant abstractions in the agent base. (I wouldn't agree fully on that as it is an area that anyone can have different opinion on what is a redudant abstraction)

Breaking views into logical modules rather than dropping them all into main.py. (I fully agree again)

These are all best practices I’d absolutely follow in production code and more as the codebase is not 100% structured robustly, and it’s great to see them highlighted here so others reading can also learn from the contrast between “demo” and “real-world” implementations.

Again, thanks for diving in this kind of feedback is exactly what makes sharing experiments valuable.

Thanks a lot, I believe too that MCP and A2A are not going anywhere anytime soon. Although both of them need more features and security hardening in order to really survive in production like or enterprise environments.

Hey all! I recently gave a workshop talk at PyCon Greece 2025 about building production-ready agent systems.

To check the workshop, I put together a demo repo: (I will add the slides too soon in my blog: https://www.petrostechchronicles.com/) https://github.com/Aherontas/Pycon_Greece_2025_Presentation_...

The idea was to show how multiple AI agents can collaborate using FastAPI + Pydantic-AI, with protocols like MCP (Model Context Protocol) and A2A (Agent-to-Agent) for safe communication and orchestration.

Features:

- Multiple agents running in containers

- MCP servers (Brave search, GitHub, filesystem, etc.) as tools

- A2A communication between services

- Minimal UI for experimentation for Tech Trend - repo analysis

I built this repo because most agent frameworks look great in isolated demos, but fall apart when you try to glue agents together into a real application. My goal was to help people experiment with these patterns and move closer to real-world use cases.

It’s not production-grade, but would love feedback, criticism, or war stories from anyone who’s tried building actual multi-agent systems. Big questions:

Do you think agent-to-agent protocols like MCP/A2A will stick?

Or will the future be mostly single powerful LLMs with plugin stacks?

Thanks — excited to hear what the HN crowd thinks!

The paper has to do about: UWB-Based Indoor Navigation in a Flexible Manufacturing System Using a Custom Quadrotor UAV

We tested whether it’s feasible to build an indoor heavy duty UAV for warehouses, and even deploy swarms. Using Marvelmind's UWB sensors(we also compared with Decawave's) on a large manufacturing floor with CNCs and lathes, we achieved reliable navigation accuracy. Blog post with details coming soon: https://www.petrostechchronicles.com/

I have the same question, do you use already an on prem RAG system?

Really glad it helped, feel free to ask anything on it

Thanks a lot, I am trying to keep up with the AI Hype so concluding all this in an article I thinks it helps everyone to catch up with MCP and ACP. Probably they are going to be the new HTTP like protocols.

Lately I have been starting playing with MCP and I started creating small use cases with it in my homelab, so I gathered some information about it and I wanted to share it with you. I hope you find it useful.

In this article I am testing and learning capabilities of MCP + ACP.

Thank you a lot! In what points do you agree most?

10 Tech Predictions for 2025 (Personal)