Returning to the basics of AI agents
Getting started with AI
at work
Selecting tasks to delegate, experimenting on a small scale, and operating safely.
- How AI works
- Trying it out at work
- Using it safely
Featured articles
Using AI output at work
Gemini's answers became "interactive diagrams" — What to decide before using them for internal explanations
Set guidelines for generated images
Google Pics is on by default: what to decide before employees start generating images
Organizing code and data governance
How far can AI be used in projects where code cannot leave premises? — The positioning of Junie Local
Verifying authentication in MCP integrations
Where MCP's roadmap still shows room to maneuver: should you connect internal systems today?
Determining operational permissions granted to AI
The era of AI writing directly to sales data — the decisions Claudeforce leaves to you
Turning concepts into design.
Development Environment↔AI Models
Access Permissions · Data · Logs
Building an environment
to use AI internally
Internal data, development environments, and access permissions. Advancing adoption concepts into concrete technical reviews.
Compare approachesOrganize implementation requirementsTest on a small scale
Read feature 
Google Workspace
administration basics
Resolving administrator uncertainties from rollout to sharing and permissions.
Pricing & rolloutSharing rulesPermissions & security
Read feature 
Looking back at
emergency vulnerability responses
PHP, certificates and Node.js. A record of the responses that overlapped on October 21–22, 2026, down to the steps we verified and the pitfalls we hit.
Checking providers' responsesInspecting certificates on devicesUpgrading multiple repositories at once
Read feature 
Improvement notebook for
growing your website
Forms, usability, and updates: tackling post-launch improvements one step at a time.
Form optimizationUI refinementsUpdates & operations
Read series