Repository navigation
docs: clarify COMA security and distribution scope - #7
Conversation
|
Welcome! 👋 Thanks for your first pull request in this repository. A maintainer will review it soon. Please make sure:
Thanks for contributing! |
There was a problem hiding this comment.
Isolated Luna review at exact PR head 102fe645c1a60aff25cbb028d1b78b73b2278056 (base 8f26cf71b6e1161c3987c5fe0192c6ba8c7a02e2).
The documentation and module metadata now describe COMA as an agent-process lifecycle layer with file-based coordination and optional starter generation, reflecting the reviewed subprocess, network, permission, writer-convention, and best-effort-cleanup boundaries. No sandbox or provider guarantee is claimed.
Validation: all 24 final CI matrix checks passed (runs 38171219007 and 38171215598). Ruff and diff checks passed; the final two-line .gitignore cleanup removes only redundant host patterns. The isolated full suite passed 301/301 with jsonschema enabled on the prior reviewed head; the author reports 301/301 after the final cleanup. Twelve dated/marketing-text assertion tests were removed while functional tests remain. Internal care logs are untracked and excluded from the reviewed wheel and sdist artifacts. Version 0.3.2 remains frozen; no release or upload was made.
Scope follow-up: generated relative schema provenance and test discovery are unchanged. Three local drift checks and the central check found no drift; configurable neutral provenance is documented for follow-up. No live provider was started, and this review is not a provider or security certification.
What changes
Validation
python -m ruff check .python -m compileall -q coma comas testspython -m pytest -ra -v: 301 passed in the clean tracked checkoutgit diff --checkNo release or package publication is included. The PyPI project named
comais unrelated; install instructions use this GitHub repository.Review focus: claim scope and EN/DE consistency, Markdown navigation/examples, tracked-file hygiene, unchanged runtime/API behavior, and frozen version 0.3.2.