Commands

Strategy

Execution plan preview

Strategy is a dry run — it shows you exactly what the system would do with your next message: intent classification, model selection, active rules and skills, and vault context that would be loaded.

How it works

See the plan before
you commit

Type /strategy followed by what you'd normally type. The system classifies intent, selects a model, loads context, and shows you the full execution plan — without executing anything.

Intent Classification

See exactly how the system interprets your message: QUESTION, EDIT, CREATE, SCAFFOLD, or REVIEW. Understand the dispatch path before committing.

Intent: CREATE | Scope: single-file | Domain: legal | Complexity: medium

Model Strategy

Which model will be used. Which rules are active. Which skills are loaded. Which commands are available. Full transparency into the AI configuration.

Model: claude-sonnet | Rules: 4 active | Skills: 2 loaded | Commands: 15

Context Preview

What vault documents would be loaded as context. What the system 'knows' for this specific request. See the RAG results before execution.

Context: 3 vault docs loaded (contract-template.md, nda-v2.md, terms.md)

No Side Effects

Pure preview — nothing is executed, no files created, no API calls made. See the plan, adjust your prompt if needed, then send for real.

Dry run complete. No changes made. Ready to execute when you are.

Transparency

Not "I hope it does the right thing."
Full visibility into the AI's decision-making.

Blind Execution
Send message and hope for the best
Wonder which model was used
No idea what context was loaded
Surprised by unexpected behavior
With /strategy
See intent classification before execution
Know exactly which model, rules, and skills are active
Preview which vault documents would be loaded as context
Adjust your prompt based on the plan, then execute with confidence

Before any important request

Look before you leap.

Preview the plan for any request you want to get exactly right.

Complex Requests

Multi-step tasks — verify the system understood your intent

First-Time Commands

New command or skill — see what it activates before running

File Operations

Creating or editing files — confirm the right model and context

Debugging Responses

AI not behaving as expected — see what rules and context are active

High-Stakes Requests

Before sending anything important — verify the full execution plan

Sensitive Operations

Legal or compliance work — confirm right model and rules are active

Curious about the plan?

In any conversation, run /strategy from the Conductor.

Get AskQP