Deep investigation — ecosystem packages, external landscape, architectural analysis, memory recall. When the answer is not in the code you already read.
Sometimes the answer is not in memory and it is not in the code you already read. That is what this is — the deep dig.
Argument: $ARGUMENTS — the question or topic to investigate.
Skip any domain that does not apply. If the question spans multiple, run all of them.
| Domain | Question it answers | How |
|---|---|---|
| Ecosystem | Does zoobzio already have this? | Run , search workspace for packages, check cross-package deps and established patterns |
/grok| Landscape | What do other people do? | Search GitHub and pkg.go.dev for community packages, assess common vs novel, check RFCs and standards |
| Architecture | Is this the right way? | Trace structural consequences, find hidden decisions in abstractions, describe concrete alternatives, check precedent |
| Memory | What do the dead remember? | Search project memories for prior context, cross-repo knowledge, pattern recurrence, what other agents asked |
Tip: ecosystem first. If zoobzio already solved this, stop. The caller does not need a landscape survey of a solved problem.
Tip: when checking precedent, grep for old implementations and read git history. Patterns that were tried and reverted are the most valuable findings.
Tip: if research produces knowledge worth preserving, write a memory before reporting. The next version of you should not have to dig this up again.