From Capable to Trustworthy: How KCP Evolved from Discovery to Governance
AI agents are getting remarkably good at doing things. They read code, traverse APIs, generate summaries, and execute multi-step plans across sprawling...
Tip #23: Write scripts that guide the agent
Script output goes straight into the agent’s context window. Design it to say what happened and what to do next — not just what ran.
Native skaper ikke verdi, men skalerer middelmådighet
Løftet om å “nå ut til flere kunder uten å løfte en finger” treffer et marked som er presset på tid, ressurser og forventninger om synlighet. Etter...
Tip #22: Use scripts for repeatable tasks
Deterministic tasks deserve deterministic scripts — not AI improvisation every session.
Tip #21: Meet OpenCode — the open source coding agent
OpenCode is an open source coding agent that works with 75+ providers — and it can run on your existing GitHub Copilot license. Fast terminal UI, built-in...
Kan kunstig intelligens virkelig ikke skape noe nytt?
«Kunstig intelligens kan ikke skape noe genuint nytt», hevdes det i en fersk kronikk i Forskersonen. Men hva om denne påstanden bygger på en forenklet...
Tip #20: Give your agents their own workspaces
Use isolated workspaces so multiple agents can work on the same repo without stepping on each other.
A peek in the Mirror with Scala
TL:DR Reading SQL ResultSets in Scala 3 is surprisingly concise — even when you want List, Option, NamedTuples, or case classes. Here’s how to do it in...
Når algoritmen trykker “kjøp”
AI handler allerede på vegne av kundene dine. Har dere designet for det?
Når kunnskap finnes, men ikke blir synlig
De fleste selskaper sitter på langt mer kompetanse enn det som faktisk blir synlig utenfor organisasjonen. Gjennom strategisk kommunikasjon, PR-arbeid og...
Tobias K. Torrissen om kodekvalitet, konferanser og AI
Hva er bra kvalitet i software? Hva er bra nok kvalitet? Endres betydningen kvalitet med AI? Tobias K. Torrissen har brukt over 25 år på å bygge, forvalte,...
Tip #19: Get a second opinion from a fresh agent
Open a second session to re-ask the same question without baggage from earlier attempts.
The Human in the Loop — at Design Time
Tim O'Reilly posted something this week about craftsmanship in the AI age. The question he was circling: how do you maintain quality standards when agents are...
The Agentic Stack: Every Layer Was Built for Humans
Something caught my attention this week. JJHub launched a version control platform built on Jujutsu instead of Git, designed specifically for agentic...
Den virkelige AI-debatten handler ikke om teknologi, men om alle som kan bli overflødige
Influencer og AI-gründer Oskar Westerlin mener nordmenn undervurderer hvor dramatisk kunstig intelligens vil endre arbeidslivet. På ett viktig punkt har han...
Tip #18: Commit before you let the agent loose
Git stash or commit first, so the agent's changes are always one git diff away from a clean state.
The agents were forgetting everything they discovered
Something felt off. I'd been using the Task tool heavily — delegating research tasks to Claude subagents, letting them explore codebases, map dependencies,...
Klarspråk i digitale tjenester
Hvordan fungerer klarspråkarbeidet i norske virksomheter i dag? I denne episoden får Oskar Seljeskog, innholdsdesigner i Bouvet, besøk av Johan Tønnesson,...
Tip #17: Keep sessions short and focused
One task per session. Context fills up — fresh starts beat accumulated corrections.
Tip #16: Explore before you edit
Use the agent as a research assistant to understand unfamiliar code before changing anything.
KCP v0.10: Pre-Invocation Discovery
An agent is about to call a knowledge server. It has no idea what that server knows, how sensitive the content is, or whether the answer will fit in its...
I haven't typed apt install in three months
Last Tuesday I needed a file-watch service on my workstation. The kind of thing that monitors a directory and triggers a reindex when something changes....
The Autonomous Agentic Web Needs a Foundation Layer
Something is being assembled right now, mostly without a name for it. Production pipelines where agents write code, run tests, and open pull requests....
AI-mode for SaaS
Etter min forrige artikkel om fremtiden til SaaS-selskaper, har jeg hatt noen interessante samtaler om hvordan man kan tenke seg at SaaS-produkter vil utvikle...
FFSBS - Flinke Folk Som Bryr Seg
Hva skal egentlig til for å lykkes med IT? Er det skinnende nye metodikker, eller kanskje en ny skinnende teknologi? Eller kanskje det er litt enklere. Les...
Tip #15: MCP servers eat your context window
Every connected server costs tokens every turn. Be selective.
Tip #14: Your CLI tools are already AI tools
gh, git, jq, curl work without MCP — the agent just calls --help.
Tip #13: Give your agent a browser
Chrome DevTools MCP lets the agent click, inspect, and test web UIs from your terminal.
Tip #12: MCP — the USB-C of AI tools
One open standard for connecting agents to external systems across all tools.