I built a small routing CLI that makes AI coding a...

# android

[Rana Ahmed]

I built a small routing CLI that makes AI coding agents
explicit about Android-specific risks before writing code.

You state observable facts about a change (persistence,
concurrency, lifecycle) and a checked-in registry maps
them to required specialist checks:

Task: brownfield / android / safety
Detected: concurrency, lifecycle, persistence, schema_migration
Tier: COMPLEX
Required:
- android-persistence-migration-readiness
- kotlin-coroutines-readiness
- protected-lifecycle-transaction-review
- android-runtime-evidence-readiness

No LLM judgment — same output every run. Install is one
uvx command, no clone needed.

Would love feedback from anyone using agents on Android
codebases — especially whether the Safety profile feels
light enough to actually adopt.

github.com/RanaAhmedHamdy/AI-Workflow