Blog
Writing on durable context, AI-assisted work, and how KOS is built.
OpenAI's Saved Prompts, Agent Builder and Evals End November 30. Keep What Took Longest to BuildOpenAI's prompt objects, Agent Builder and Evals end on November 30. Move the prompt, the workflow and the test cases into files you own before they go.Build a One-Person AI Video Studio with KOS and Kraven Video StudioDesign an AI agent workflow from script to publishing with KOS and Kraven Video Studio, with recording as the target routine manual step.Every Tool Wants to Be Your Agents' Shared Brain. Decide Where Yours LivesNotion and others now pitch a shared brain for every AI agent. Decide which context belongs in a hosted workspace and which belongs in files you control.From Recording to Release: KOS and Kraven Video StudioSee how KOS gives AI-assisted video production reusable context, clear handoffs, and human review, with a full production-team example using Kraven Video Studio.Before You Connect an AI Agent, Sweep Your Folders for SecretsA CERT-In advisory shows how exposed credentials enabled unauthorized agent access. Sweep notes, exports and project folders for secrets before connecting an AI agent.Gemini Gems Are Becoming Skills. Keep Your AI Instructions Somewhere You OwnGemini Gems are moving to skills and older models keep retiring. Keep your best AI instructions as plain files you own, with thin adapters for each tool.More AI Code, Same Shipped Work? The Review BottleneckA study of coding agents found more code and longer reviews, without a statistically significant rise in completed work. Here is a practical way to make AI changes easier to review.Need Help Putting KOS to Work? Explore KRAVEN AI ConsultingGet practical help applying KOS to your work through KRAVEN AI consulting, personal sessions, team workshops, or a focused business action plan.Before You Give AI Access to Your Files, Set These BoundariesDefine what an AI agent can read, change, and publish with a practical access checklist, scoped permissions, and a safe verification exercise.Stop Using Your AI Memory as a Daily LogKeep project memory useful by separating durable decisions from daily activity, current handoffs, and completed change history.When Your Docs Disagree, Who Wins? A Source-of-Truth Hierarchy for AIResolve conflicting project context by separating approved intent, verified implementation, and current progress before AI makes the next change.One Router, Many Agents: Keep Your AI Instructions from DriftingGive multiple AI tools one maintained set of project rules, with a compact instruction router, thin adapters, and checks that expose context drift.The Second Brain Maintenance Model: Keep Useful Knowledge CurrentBuild a manageable knowledge review workflow: let AI propose filing, links, and updates while you retain control of decisions, evidence, and approval.Stop Re-Explaining Your Project to AI: A Practical Handoff WorkflowKeep decisions, current progress, and next actions clear with a practical KOS workflow using memory.md and handoff.md across AI sessions.The Kraven AI Framework: Turn AI Conversations Into Work You Can Build OnLearn the six phases of the Kraven AI Framework, with practical examples for planning, building, checking, and preserving your work with KOS.From AI to Superintelligence: What Actually Has to ChangeA practical look at the difference between useful AI and superintelligence, and what teams can do today without mistaking context for capability.Make Obsidian More Useful with KOS CommunityKOS Starter Kit v1.2.0 brings project memory, handoffs, and shared AI context to an Obsidian-compatible workspace. Here is how to put it to work.
No articles found
Try another title or topic.
More articles appear here as they are published.