Commands

Installs the pinned Node.js dependencies from package-lock.json and points Git at the repository hooks in .githooks.

Source: package.json · Verified by: ci:Validate and build, ci:End-to-end (subpath preview)

Creates the project virtual environment and installs the pinned Python test tooling.

Source: requirements-dev.txt · Verified by: ci:Python tests

Runs Tailwind in watch mode next to zola serve for local authoring with live reload.

Source: scripts/dev.sh · Verified by: pytest:test_dev_server_serves_the_site

Builds the production site and serves it under the same repository subpath GitHub Pages uses.

Source: scripts/preview.sh · Verified by: ci:End-to-end (subpath preview)

Copies the vendored browser bundles, compiles Tailwind and runs zola build into public/.

Source: package.json · Verified by: pytest:test_build_produces_the_site

Runs every static gate: toolchain, assets, JavaScript syntax, translation parity, content standards, command registry, Zola check and build, HTML and link checks.

Source: scripts/validate.sh · Verified by: ci:Validate and build

Runs the pytest suite for the validators, exporters and the command registry.

Source: tests/python · Verified by: ci:Python tests

Runs the Playwright browser suite against a local production build served under the Pages subpath.

Source: tests/site.spec.js · Verified by: ci:End-to-end (subpath preview)

Checks that commit subjects follow Conventional Commits; the same script backs the commit-msg hook.

Source: scripts/lint-commits.sh · Verified by: ci:Conventional commits, pytest:test_lint_commits_accepts_and_rejects

Checks content standards (evidence fields, SEO descriptions, banned promotional phrasing) and that every documented command is registered and linked.

Source: scripts/validate-content.py · Verified by: ci:Validate and build, pytest:test_content_passes_on_repository

Checks that nothing on the site scrolls inside the page: no scrolling overflow in the stylesheets or templates and no minimum width inside a figure or table.

Source: scripts/validate-layout.py · Verified by: ci:Validate and build, pytest:test_layout_passes_on_repository

Checks the interactive model registry against docs/MODELS.md: bilingual fields, claims and sources that resolve, an epistemic class for every part of the model, assumptions, limits, rules, an implementation and its tests.

Source: scripts/validate-models.py · Verified by: ci:Validate and build, pytest:test_registry_passes_on_the_repository

Checks the evidence ledger: every reference appraised, claims graded no higher than their sources, review intervals and dates, coverage of every research note and advice entry, and the glossary.

Source: scripts/validate-claims.py · Verified by: ci:Validate and build, pytest:test_ledger_passes_on_repository

Checks the audience taxonomy: the six lenses and the audiences that map onto them, bilingual labels, starting points that exist, disclosure slots a component renders, and an epistemic vocabulary of exactly six words.

Source: scripts/validate-audiences.py · Verified by: ci:Validate and build, pytest:test_audiences_pass_on_repository

Validates the reference registry offline; with --online it compares every DOI, title and year with Crossref.

Source: scripts/check-references.py · Verified by: ci:References (Crossref), pytest:test_references_pass_offline_on_repository

Runs the full browser suite against the live GitHub Pages URL (BASE_URL overrides it).

Source: package.json · Verified by: ci:Verify production

Checks that every page, feed, preview image and asset of the live site answers with the expected status.

Source: scripts/smoke-production.sh · Verified by: ci:Verify production

Lists the most recent runs of the Pages pipeline with their status.

Source: .github/workflows/pages.yml · Verified by: pytest:test_gh_run_list_reads_the_pipeline

Prints the next version and release notes derived from the commits since the last tag, without publishing.

Source: scripts/release.sh · Verified by: ci:Release, pytest:test_release_dry_run_computes_a_version

Exports research notes, advice, references and the command registry as the versioned JSON API under api/v1.

Source: scripts/export-api.py · Verified by: ci:Validate and build, pytest:test_export_api_writes_the_contract

Runs the Rust unit tests and the contract and CLI tests against the exported API.

Source: Cargo.toml · Verified by: ci:Rust

Lints the Rust workspace and treats every warning as an error.

Source: Cargo.toml · Verified by: ci:Rust

Checks that the Rust code is formatted with the pinned rustfmt.

Source: rust-toolchain.toml · Verified by: ci:Rust

Runs the caas command-line client: list or show advice, list research, list claims due for review, show a claim or glossary term, check the library, print the endpoint response.

Source: crates/caas-cli/src/main.rs · Verified by: ci:Rust

Begins a supervised task with a declared scope of paths the work may touch.

Source: .ai/repo/workflows/task-lifecycle.md · Verified by: pytest:test_majordomus_commands_are_available

Reports whether the active task is consistent with policy, scope and recorded state (read-only).

Source: .ai/repo/workflows/task-lifecycle.md · Verified by: pytest:test_majordomus_commands_are_available

Records a short progress note for the active task.

Source: .ai/repo/workflows/continuity.md · Verified by: pytest:test_majordomus_commands_are_available

Evaluates the finish contract and closes the task only if the verification command passes; --check runs in the pre-push hook.

Source: .githooks/pre-push · Verified by: pytest:test_majordomus_commands_are_available

Reports drift between the recorded task, its scope, the policy and the generated projections.

Source: .ai/repo/policy.yaml · Verified by: pytest:test_majordomus_commands_are_available

Checks that the supervision layer is healthy and wired; runs in the pre-commit hook and in CI.

Source: .githooks/pre-commit · Verified by: ci:Majordomus supervision

Reads back the append-only ledger of task starts, checkpoints, handovers and finishes.

Source: .ai/README.md · Verified by: pytest:test_majordomus_commands_are_available

Writes a continuation record with the objective, current state and next action.

Source: .ai/repo/workflows/continuity.md · Verified by: pytest:test_majordomus_commands_are_available

Records a durable decision with its rationale, rejected alternatives and evidence.

Source: .ai/README.md · Verified by: pytest:test_majordomus_commands_are_available

Exports the curated case-study snapshot (data/case_study.json) from the Majordomus ledger, Git and GitHub; --check validates the snapshot without local state or network.

Source: scripts/export-case-study.py · Verified by: pytest:test_case_study_check_passes_on_repository, pytest:test_case_study_export_is_deterministic

Lists the effective rule set (vendored baseline plus project rules) with class and the commands that enforce each rule (read-only).

Source: .ai/repo/rules/README.md · Verified by: pytest:test_majordomus_read_only_commands_used_by_the_case_study

Lists architecture decision records with their status; acceptance stays a person's act (read-only).

Source: .ai/repo/adrs/README.md · Verified by: pytest:test_majordomus_read_only_commands_used_by_the_case_study

Re-renders the poster master, the social preview cards and the icon set from the sources in artwork/.

Source: scripts/render-artwork.sh · Verified by: pytest:test_render_artwork_check_mode

Renders a 1200×630 social preview for every content page and language from artwork/og.html; with --check, verifies without a browser that every page has an up-to-date preview.

Source: scripts/render-social.py · Verified by: ci:Validate and build, pytest:test_social_check_reports_a_missing_preview