How I write, review and ship code across every repo under @Abdel-Abdelaal. Two jobs:
- The handbook in
standards/: the written rules for commits, PRs, issues, code review, Python code, repo layout, versioning, decisions and AI use. - Default community files. GitHub uses the issue templates, PR template,
CONTRIBUTING.mdandSECURITY.mdin this special.githubrepo for any of my repos that doesn't have its own. Every repo gets them without copying.
Most rules here are enforced by a tool or a CI check, not by memory. Each standard says which. The CI checks live in platform-workflows.
| Standard | Covers | Built on |
|---|---|---|
| Commits | commit messages, PR titles | Conventional Commits 1.0 |
| Pull requests | size, description, merge flow | Google eng-practices |
| Code review | what reviewers look for, comment format | Google eng-practices, Conventional Comments |
| Issues | tickets, labels, acceptance criteria | GitHub issue forms, labels as code |
| Python | naming, comments, docstrings, types, duplication | PEP 8, Google Python Style Guide, Ruff |
| Repo structure | layout and required files per repo type | PyPA src layout, Standard Readme |
| Versioning | releases, tags, changelogs | SemVer 2.0, Keep a Changelog |
| Decisions | recording why we chose something | MADR |
| AI | where AI runs, API keys, what it may do | AGENTS.md, opencode |
| Languages | which tool checks which language, for every repo automatically | super-linter, CodeQL, language toolchains |
| Governance | how everything is re-checked and kept current, settings as code | Terraform, Dependabot, OpenSSF Scorecard |
Open a PR here. A standard changes when the tool or check that enforces it changes in the same release, so the written rule and the enforced rule never drift apart.
Same-repository PRs receive an isolated MiMo 2.6 Flash review; fork and Dependabot PRs skip. Findings are non-blocking summaries and neutral notice annotations. Weekly audits maintain one health issue. A bounded, tested draft fix PR is possible only after the automation App is configured; otherwise that step skips explicitly. No AI check is required for merging, and no AI workflow can merge or approve. See the isolation design.
Maintenance callers now pin platform v1.9.1 (7b7bea4e50c6ee1b1dcf0d318e4504655678b0b8).
Signed AI evidence and measured runs appear in the
public AI observatory. Frozen hook checks
run where a hook configuration is supported; native Dependabot maintains baseline
hook revisions. Stale/release mutations remain explicitly gated on canonical
labels and maintenance App setup. The Python audit includes development tooling.