Skip to content
View kimasplund's full-sized avatar

Block or report kimasplund

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
kimasplund/README.md

Kim Asplund

AI agent engineer in Vaasa, Finland. I build agents that run real workflows in production, and the tooling that shows whether they behave: evaluations, tracing, and guardrails enforced in code rather than in prompts.

asplund.kim

In production

Aino runs the EU B2B sales of a product line end to end: prospecting, outreach and customer correspondence in four languages, the CRM, and orders booked into Odoo through to invoice. It's built on Hermes Agent and Claude. The guardrails live in code: a banned-claims preflight across 18 languages, a tool-call guard, approval gates before a new market goes live, and audit trails for opt-outs and customer promises. The code is private because it runs a live business.

Evaluation and observability

  • thinkbetter-skills: 15 reasoning skills for coding agents, shipped with skill-gauntlet, a pre-registered blind A/B harness that decides whether a revised skill is actually better. Every scorecard is published, including the revisions that failed and were reverted.
  • LLM-Context-Trace-Library: time-travel debugging for multi-agent workflows. Record a run, replay it to any step, and diff two runs to find where they diverged. Integrations for LangChain, CrewAI, AutoGen, OpenAI Agents, PydanticAI, Semantic Kernel, DSPy and LlamaIndex.

Tools for agents

  • mcp-pypi: MCP server that lets agents search PyPI and check packages and whole dependency trees against the OSV vulnerability database before they suggest anything. Available on PyPI.
  • nokka and kyyhky: label-printer drivers that speak the printers' own protocols (Datamax DPL, Brother QL raster), each with an agent skill so a coding agent can drive real hardware without wasting labels.

Systems

Background

Before moving into AI in 2020, I spent more than 20 years in technical service engineering (medical devices, IT infrastructure) and running my own businesses, plus ten years as a firefighter and ambulance medic. That's why my work centres on systems that must not fail quietly.

Python, Rust, TypeScript. Swedish (native), English, Finnish.

Pinned Loading

  1. thinkbetter-skills thinkbetter-skills Public

    Reasoning skills for coding agents: a scored router, nine reasoning patterns, premortem, option synthesis, and a blind A/B skill-evaluation harness

    Python

  2. LLM-Context-Trace-Library LLM-Context-Trace-Library Public

    Time-travel debugging for multi-agent LLM workflows. Replay execution to any point, compare runs, find bottlenecks.

    Python 1

  3. mcp-pypi mcp-pypi Public

    MCP server for PyPI: search packages and audit dependency trees against OSV vulnerability data before an agent suggests them.

    Python 3 2

  4. kimsfinance kimsfinance Public

    Fast financial charting for Python: candlesticks and 32 indicators, with optional Rust and GPU acceleration.

    Rust

  5. rust-p2p-usb rust-p2p-usb Public

    Use a USB device on another machine anywhere on the internet, peer to peer over QUIC with Iroh. Alpha.

    Rust 1

  6. gnome-oled-shield gnome-oled-shield Public

    GNOME Shell extension that protects OLED screens from burn-in: pixel shift, pixel refresh, dimming of static UI.

    JavaScript 32 1