No blind writes
Read the exact block before it is written so a shared instruction file only ever receives context you have seen and approved.
Auto-Inject
Preview and refresh
See the exact brief before it lands, and refresh it whenever the vault moves on.
Overview
What Preview and refresh does
Preview shows the exact managed block Auto-Inject will write before you enable or refresh a target. Instead of trusting the feature blind, you read the real text an agent will see and confirm that only the intended preferences, decisions, and pinned memory are included. It is the safest way to verify an injection, especially before pointing it at a shared instruction file.
Vault memory changes as you capture and classify, so a brief written last week can drift. Refresh regenerates the block on demand from current memory, and the preview surfaces problems first — stale classification or missing imports show up as thin or empty context before an agent ever relies on it. You fix the vault, refresh, and the brief catches up.
PreviewRefresh
Why it matters
The payoff for your AI memory
Refresh on demand
Regenerate the brief whenever the vault changes, keeping injected memory in step with your latest captures and classifications.
Spot gaps early
A thin or empty preview reveals stale topics or missing imports, so you notice a coverage hole before an agent starts from incomplete context.
How it works
From first launch to reusable memory
- 1
Open a target in the Auto-Inject tab and click Preview.
- 2
Read the managed block and confirm only the memory you intend is present.
- 3
Fix any stale or missing context in the vault, then refresh to rewrite the block.
FAQ
Common questions
Can I see what auto-inject will write before enabling it?
Yes. Preview renders the exact managed block for a target before you enable or refresh it, so you can confirm the injected memory before anything is written to an instruction file.
How do I update the injected context after my vault changes?
Use Refresh to regenerate the block from current vault memory. The preview also flags stale classification or import gaps, so you can repair the vault first and then refresh.
Related
More Auto-Inject features
Auto-Inject Memory
Start new agent sessions with a fresh vault brief so important memory reaches the CLI before you repeat context.
Learn moreMinimal, Standard, Rich, and Max profiles
Choose how much memory is written into agent startup files based on the work and model context budget.
Learn moreProject-scoped injection
Auto-Inject can scope memory to the current project so agents start with context that matches the workspace.
Learn moreLocal memory, shared everywhere
Give every AI tool the same memory.
Start free, import real conversations, and reuse your memory across every AI agent you already use.