Evidence: issue-185 verification runs

Captured 2026-08-09 in the work item's own checkout (cli/, via uv run). Nothing redacted — no tokens, hostnames or personal data appear in these outputs.

Red → green

The suite was written against the design before the parser handled bold-wrapped markers; the red run is the TDD record, the fix followed, then green.

Red (first run of the new suite)

FAILED tests/test_graph_contribution.py::test_parse_goal_tolerates_markdown_decoration
1 failed, 5 passed in 0.13s

(Two later reds in the same session — the thread-path authorization mismatch between bare API logins and @-prefixed allowlists — are recorded in the execution log; both fixed in goal.py, never by weakening a test.)

Green — the new suite

$ uv run pytest tests/test_graph_contribution.py -q
33 passed in 1.48s

Full suite

$ uv run pytest -q
1558 passed, 1 skipped in 47.63s

The one skip pre-exists this work item (environment-dependent, unrelated).

Review round human-1 (PR #187): the uninitialized repository

Seven tests added for R6 (defaults with no .the-loop/; the walk in a never-adopted git repo with the spec tree excluded — git check-ignore + clean status --porcelain; adopted repos untouched; publish-artifact posting only where the thread is the surface), then the full suite re-run:

$ uv run pytest tests/test_graph_contribution.py -q
40 passed in 3.03s

$ uv run pytest -q
1565 passed, 1 skipped in 49.62s

Lint and type checks

$ uv run ruff check .
All checks passed!

$ uv run pyright
0 errors, 0 warnings, 0 informations

$ npx markdownlint-cli2 <the 16 changed/added markdown files>
Summary: 0 issues in 0 files