Search Unreal Engine documentation

Your AI client's engine knowledge is generic and can be outdated or wrong for your engine version. The research tools ground it in curated Unreal Engine documentation, known patterns, and your project's own logs — and they're all free.

Search the documentation

When you want the answer grounded in the engine docs, say so:

Using search_docs_code: how does FTickFunction ordering work during level load?

Your client gets back ranked excerpts with source paths, and answers with those in hand instead of from memory.

Check an approach against known patterns

Before committing to a design, check it:

Using search_patterns: I'm spawning actors in Tick to trail the player. Is that a known anti-pattern?

You get the recommended engine approach, the anti-pattern to avoid, and the rationale behind both.

Search your project's logs

After analyzing logs, the ingested events are searchable by time, severity, subsystem, and build:

Using search_logs_temporal: find PackageName warnings in the last week, severity Warning.

Investigate one incident

For a single stubborn problem, let the engine combine logs, relations, and docs into one investigation:

Using investigate_incident: investigate the intermittent load-hitch from yesterday's build, subsystem Loading.

The log, relations, and incident tools need team/project context — your AI client wires it via setup_spaghetti before first use (see Teams and projects). The documentation and pattern searches work everywhere.

Verify

• Answers cite engine documentation or your project's logs rather than answering from generic memory.

Next steps

• Research tools reference — every tool, its filters, and outputs