Decisions you have to make about AI agents.
The calls without an obvious answer, and what actually changes them. Every page names all the real options, including the ones that are not Wire.
Every answer, guide, and post grouped by topic is in Learn.
Keep the retrieval, and fine-tune only what retrieval cannot fix.
We are about to commit another quarter to our retrieval pipeline. Is RAG still worth it, or should we fine-tune?
Retrieval
Usually a better retrieval pipeline first.
Our RAG pilot keeps missing links between records. Knowledge graph, or better retrieval?
Retrieval
Pick by where the data lives, not by which protocol looks strongest.
Our internal knowledge sits in a wiki, a shared drive, and a CRM. How do we get agents reading it?
Internal data
Re-explaining is a retrieval failure, not a memory failure.
Is AGENTS.md enough, or does my codebase context need somewhere else to live?
Coding agents
Disconnect servers before you optimize them.
Our five connected MCP servers eat the context window before any work starts. What do we cut?
AI costs
Freshness is a refresh-schedule problem, not a placement problem.
Our documents change daily and the index is always behind. Process at upload, or at query time?
Retrieval
Scope access per task, not per agent.
Our agents hold production credentials. How do we scope them down without breaking them?
Security
Capture the decisions, not the documentation.
A senior engineer is leaving in three weeks. What's actually worth capturing?
Knowledge retention
Hit a problem we haven't answered?
Tell us what you're stuck on. We answer the ones we can answer honestly.
Send us the question