The SDK is primarily a library. The CLI entry point exists so a new user
typing python -m minerva gets a useful response rather than silence.
python -m minerva # banner + cheat-sheet
python -m minerva --help # argparse help with usage epilog
python -m minerva --version # → minerva-sdk X.Y.Z
The cheat-sheet covers:
- Credential resolution (constructor args vs env vars)
- Quick examples for
mc.api.resolve, mc.api.enrich, mc.api.call, mc.usage.tallies
- The two resource namespaces (
mc.api, mc.usage)
- Where to dig deeper (
help(Minerva) in REPL)
Env vars the SDK reads
| Variable | Purpose |
|---|
MINERVA_API_KEY | Your x-api-key. Required if you don’t pass api_key= to the constructor. |
Just the one. The SDK never reads any other env var.
For per-method docs, REPL is faster than this docs site —
help(mc.api.enrich) shows the exact signature, args, returns, raises, and
an example for one method.