Second release. Rebrand to moor and quality-of-life improvements.

What's new

• Rebranded to moor from agentic-python-template. Old URLs auto-redirect, existing clones keep working.
• Lower pylint default: 9.5 instead of 10 in the setup wizard. Pylint 10 is the score of a blank file; real code lands at 8.x to 9.x and felt punitive on first run. 10 is still available as an opt-in.
• scripts/new-service.sh: one-command helper to scaffold an additional service from _service-template/ after init. No more manual cp + sed.
• Sharper README: etymology blockquote, "Why moor exists" section above the tooling table, explicit "this is for multi-service AI agent projects" framing with pointers to alternatives for single-service users.

See CHANGELOG.md for the full v1.1.0 entry.

First stable release of the agentic Python template.

What's in v1.0.0

Setup

• One-shot python3 init.py wizard with opt-in workflow toggles (TDD, DDD, branch protection, GitHub Actions, docstring style).
• UV-based per-service virtualenvs with direnv auto-activation.

Code quality

• Strict pre-commit: ruff (with D pydocstyle rules + NumPy convention by default), mypy strict, pylint with duplicate-code detection.
• GitHub Actions CI with a per-service test matrix.

Document Driven Design

• 5-state design-doc lifecycle (draft → ready → in-progress → shipped → superseded).
• Auto-generated solution docs when a design ships.
• Append-only ADRs for non-obvious decisions during implementation.

Three layers of AI guardrails for Claude Code

• Mechanical (always-on): Claude Code hooks for pre-edit reuse reminder, stop-time /simplify nudge, and periodic review prompt.
• Deterministic (commit-time): pylint similarities flags 6+ identical lines across files.
• Semantic (opt-in): /review slash command orchestrates three read-only review agents (dry-reviewer, simplicity-reviewer, docs-sync) in parallel.

Apache 2.0 licensed. See CHANGELOG.md for the full entry.