Evidence — lint, format, typecheck, config validation, docs parity (T12)

The make targets CI runs, on the tree this pull request proposes. hooks.preCommit / hooks.prePush are [lint, typecheck, unit-test] and pre-commit runs the same tools, so this is the CI-parity check the-loop's own reference/tooling.md requires.

Raw output

### make lint
uv run ruff check cli hooks
All checks passed!
npx --yes markdownlint-cli2@0.18.1 "**/*.md"
markdownlint-cli2 v0.18.1 (markdownlint v0.38.0)
Finding: **/*.md !**/node_modules/** !cli/node_modules/** !**/.venv/** !docs/.vitepress/dist/** !docs/.vitepress/cache/** !docs/operating-model/reference/**
Linting: 775 file(s)
Summary: 0 error(s)

### make format-check
uv run ruff format --check cli hooks
235 files already formatted

### make typecheck
uv run pyright cli
0 errors, 0 warnings, 0 informations

### make validate
uv run python scripts/validate_config.py
VALID   .the-loop/harness-config.yaml
VALID   skills/the-loop/templates/harness-config.yaml
VALID   cli/the_loop/harness-config.default.yaml
VALID   .the-loop/collaborators.yaml
VALID   skills/the-loop/templates/collaborators.yaml
VALID   .the-loop/cli-config.yaml
VALID   skills/the-loop/templates/cli-config.yaml

### docs<->schema parity
........                                                                 [100%]
8 passed in 0.09s

What each gate proves for this work item

Gate	Why it is here
ruff check	the new module-level constants, the parse function and the require_branch plumbing carry no unused import, shadow or undefined name
markdownlint (775 files)	the spec chain, the decision and every documentation page edited here pass the project's markdown rules — including the two long tables in routing-options.md and the capability page
ruff format --check	the tree matches the formatter, so pre-commit will not rewrite it under a reviewer
pyright	TmuxConfig.session_per_pr is a str and every construction site was updated to pass one. Two test call sites were passing False positionally-by-keyword; they now say "never", which is what they meant. The registry's session_for(session_per_pr=...) stays a boolean on purpose — policy is the caller's — and typechecks as one
validate_config.py	this repository's own .the-loop/cli-config.yaml (now stating sessionPerPr: cross-repository explicitly) and the shipped template both validate against the changed schema
test_docs_parity.py	P3/P4/P5 — the documented tmux.sessionPerPr option still exists in the schema, every schema leaf is still documented, and the option's heading still states Type and Default. This is the gate that would have caught a schema change shipped without a doc change
test_config_schema_parity.py	.the-loop/cli-config.schema.json and the packaged copy under cli/the_loop/schemas/ are byte-identical after the edit

No secret, token, hostname or personal datum appears above: it is tool banners, file counts and repository-relative paths.