Python Service: Bootstrap Completeness

Repository: mem-labs/mem

activeUpdated Feb 19, 2026, 8:40 PM

Name

Status

Markdown

New deployable Python services must include baseline operability scaffolding.

Rationale:
- Engineers should be able to install, run, test, lint, and type-check new services without tribal knowledge.

Scope:
- PR review policy for new deployable service roots under `domains/python/**`.

Definitions:
- Service root: a newly added service directory containing runtime entrypoint (`app/main.py`) and/or service Dockerfile.

Trigger:
- A PR adds a new service root in scope.

Approval checks:
1) Service root `README.md` includes exact local commands for install/start/test/lint/type-check.
2) Service includes baseline scripts:
   - `script/start`
   - `script/install`
   - `script/test`
   - `script/lint`
   - `script/type_check`
3) Baseline scripts are runnable (executable + shebang, or explicit `bash script/<name>` invocation documented in README).
4) Service includes baseline config:
   - `pyproject.toml`
   - one dependency lock strategy (for example `uv.lock`, `poetry.lock`, or pinned requirements lock pair)
   - lint/type-check configuration (in `pyproject.toml` or dedicated config files).

Evidence:
- Diff shows required files and runnable command paths.

Exception path:
- Allow deviation only when all are true:
1) PR description includes `Policy Exception: python-service-bootstrap-completeness`.
2) PR explains which baseline element is deferred and why risk is acceptable.
3) PR includes tracked follow-up owner and due date.

Decision:
- Return NOT APPROVED when trigger is met and baseline checks fail without a valid exception.

Cancel