.github — Lentago Labs org defaults and fleet governance

This is the Lentago Labs organization's special .github repository. It serves two roles: GitHub reads org-level defaults from here, and it houses the settings-as-code tooling that governs the rest of the fleet.

Org profile

profile/README.md renders as the organization profile page at github.com/lentago. profile/assets/banner.svg is its header.

Community-health defaults

CONTRIBUTING.md, SECURITY.md, CODE_OF_CONDUCT.md, and LICENSE live at the repo root and apply org-wide to any Lentago Labs repo that doesn't define its own. Issue and PR templates are not yet set.

Fleet governance

fleet-ops/

Settings-as-code for the whole fleet.

• fleet-apply.sh — drift checker and enforcer. Verifies (and with --apply fixes) per-repo branch rulesets, required status checks, merge-button options (squash-only, auto-merge, delete-branch-on-merge), and the lentago+claude topic spine across all non-archived org repos.
• repo-ruleset.json — per-repo branch-protection ruleset template (PR required, squash-only, no force-push/deletion).
• labels.json — the Tidewater issue-label palette, applied with --apply-labels. Aligns color and description on the labels it names and never deletes the ones it doesn't.
• org-ruleset.json — org-level ruleset definition.
• required-checks.json — per-repo map of the status checks that must pass before merge.

fleet-reports/

Periodically regenerated reports published in this repo.

• fleet-report.md — the weekly fleet report: open issues by repo, a 30-day merge/close activity snapshot, and an instruction-as-code language census. Auto- refreshed every Monday by a GitHub Actions workflow.
• incidents.md — the public incident register, linking to post-mortems under fleet-reports/incidents/. New reports are harvested from session transcripts via the local /incident-digest playbook.

metrics/

• generate-fleet-reports.py — generates both fleet-report.md and incidents.md.
• language-census.md — a periodic cloc-based language breakdown across all in-scope org repos.

ci/

• validate.py — the check that gates PRs here. Asserts that fleet-ops/*.json match the shape fleet-apply.sh consumes, that brand/generated/ is reproducible from brand/fleet.json rather than hand-edited, that the report generator's classifiers route known paths correctly, that relative markdown links resolve, and that fleet-reports/incidents.md is reproducible from its sources rather than hand-edited. Run it the way CI does: python3 ci/validate.py.

brand/

Brand assets, and the generator that turns them into per-repo identity.

• avatars/ — the Lentago Labs mark in SVG and PNG variants (square and circular, teal and limestone colourways).
• marks/ — the 64-grid genus marks, one per fleet system, plus the lentago blossom that every other repo falls back to.
• fleet.json + generate.py — per-repo identity as config: emits each repo's README banner, badge row, and 1280×640 social-preview card into brand/generated/. render.sh rasterizes the cards with the real brand typefaces. Generated output is CI-enforced against hand-editing.

Social previews are the one surface with no API — they're uploaded per repo under Settings → General → Social preview.

---

Fleet CI conventions and the reusable workflows other repos call live in shared-workflows, not here. The workflows under .github/workflows/ are this repo's own — they gate its PRs and refresh the fleet reports.

---

Built in collaboration with Claude (Anthropic) — directed and reviewed by an infrastructure operator, not a software engineer.