base120 SDK
Python SDK for HUMMBL Base120 mental models. Stdlib-only — no third-party runtime dependencies. Includes Engine.select() for model recommendation, Krineia append-only audit ledger, MCP server (JSON-RPC 2.0), and CLI.
Doctrine terms
This report references these HUMMBL doctrine terms:
100.0
Category Breakdown
Key Modules
| Module | Purpose | Status |
|---|---|---|
engine.py |
Keyword-overlap Engine.select() for model recommendation | Pass |
models.py |
120 operator registry with metadata (6 families, P/IN/CO/DE/RE/SY) | Pass |
ledger.py |
Krineia append-only audit log — OperatorTuple (id/time/state/drift) | Pass |
mcp_server.py |
JSON-RPC 2.0 MCP server — 6 tools (get, list, families, prompt, select, record) | Pass |
cli.py |
CLI for list, get, prompt, and families commands | Pass |
Findings
No lint, security, complexity, dead code, or duplication findings. All 157 tests pass in the current package verification. Clean ruff output.
All 5 modules use only Python stdlib. Use the source checkout until PyPI publication is verified; the SDK has no runtime third-party dependencies.
Append-only JSONL ledger with exactly 3 public methods:
append() and cut(). Operator tuples are
immutable after writing — no delete, no mutation.
Want an audit like this for your codebase? Free for open-source repos.
Request Free Audit