The Context Log
Your logbook for smarter projects. Insights on local-first AI project management, agents at work, and turning chaos into clarity — built on the principles of Context-Driven Project Management.


Which AI model plans a project best? 14 models, three projects and one change halfway through
14 AI models planned the same projects and got a change halfway through. Three deliver plans you can build on. But how well a model plans says little about how well it replans.


Spotting schedule delays in construction projects early: critical path, float and snapshots
A delay announces itself weeks ahead, in an email, a shrinking float, a predecessor that slipped. Three questions to the schedule surface it in time.


TensorPM, Claude Code and Codex: Putting the project plan first
Understand, Steer, Execute: Clarify goals, assess the plan and current situation, and get work done. TensorPM connects these activities through a shared project context.

Codex, Claude and Co.: The Wrong Approach
ChatGPT Work and Claude Cowork deliver finished documents for review. Real collaboration begins when person and AI work on the same living artifact.

The Model Matters Less Than the Context: Introducing ProjectBench
We built a benchmark that runs real project-management work inside TensorPM and scores each plan on one thing: would the owner actually be better off building from it? The first finding is the whole thesis of Context-Driven Project Management. The project context decides plan quality far more than which model you run.


AI Agents for Project Management 2026: A Practical Comparison
TensorPM, OpenClaw, Hermes Agent, Pi, and DeepSeek Harness compared: project steering, messenger automation, self-improving workflows, or technical execution?