feat: Implement github_create_issue in write_tools.py + test

[mabry1985]

I'll start by reading the canonical instructions and the files I need to modify.I'll start by reading the canonical instructions and the files I need to modify.Let me look at read_tools.py for the implemented-tool conventions and the existing tests for style.Let me read the existing test files to match conventions.Now I have everything I need. Let me implement the changes. First, update the import in write_tools.py:Now replace the stub body of github_create_issue:Let me check the test config for async support before writing the test.asyncio_mode = "auto" is set and pytest-asyncio is a dev dep, so async test functions run directly. The tool calls run_gh/check_gh_error via the ghplugin.write_tools module namespace, so I'll patch there. Now I'll write the test:Let me verify my work by running the lint and test gate.Ruff passes. The pytest failures are environmental — the homebrew pytest lacks pytest-asyncio, so asyncio_mode=auto is unrecognized (the 6 sync tests pass). Let me confirm the new tests actually pass with the real dev deps in a clean venv.All 11 tests pass (6 existing + 5 new) and ruff check/format both pass. Let me clean up the temp venv.Done. Here's what I implemented:

write_tools.py

• Updated the module-top import to from .gh_cli import bad_repo, check_gh_error, run_gh.
• Replaced the github_create_issue stub body with a real gh issue create invocation:
  • Validates repo via bad_repo() first (short-circuits).
  • Builds ["issue", "create", "--repo", repo, "--title", title, "--body", body].
  • Splits labels on comma, strips each, skips blanks, and appends a --label <name> pair per label.
  • Runs via run_gh, checks failures via check_gh_error, and returns out.strip() (the issue URL).
  • Replaced the TODO(team) docstring line with Returns the new issue URL. (kept it a plain string literal per §6).
  • github_comment and github_create_pr were left untouched.

tests/test_write_tools.py (new)

Mocks run_gh in the ghplugin.write_tools namespace (where the tool resolves it) and asserts:

• success returns the stripped issue URL, with the correct gh issue create argv (repo/title/body);
• no labels → no --label flag;
• comma-separated labels split into separate --label flags, stripped, blanks dropped;
• invalid repo returns the bad_repo error and never shells out;
• a nonzero gh exit returns the check_gh_error string.

Verification

• ruff check . → all checks passed; ruff format --check . → 8 files already formatted.
• pytest -q in a clean 3.12 venv with requirements-dev.txt → 11 passed (6 existing + 5 new). (The system pytest lacked pytest-asyncio, hence the need for the venv to honor asyncio_mode=auto.)

All acceptance criteria are satisfied.

[protoquinn]

👀 Quinn is reviewing — verdict (PASS / WARN / FAIL) + findings to follow.

[protoquinn]

QA Audit — PR #2 | feat: Implement github_create_issue in write_tools.py + test

VERDICT: WARN (CI test still queued — provisional; final verdict on terminal CI)

---

CI Status

• test: queued

Diff Review

• write_tools.py: Replaced stub with gh issue create impl. Imports check_gh_error + run_gh; validates repo via bad_repo(), builds argv, returns issue URL or error. Clean, follows existing patterns.
• tests/test_write_tools.py (new): 5 async tests — success path, no-labels, comma labels with blanks, bad-repo short-circuit, gh error propagation. All patch in correct namespace (ghplugin.write_tools).

Observations

• LOW: labels.split(",") would AttributeError on explicit labels=None (default "" makes this unlikely). Consider (labels or "").split(",").
• LOW: body=None would produce --body None in argv. Same rarity — default "" covers normal calls.

No blocking findings. Implementation is sound, tests are thorough.

— Quinn, QA Engineer

[protoquinn]

Submitted COMMENT review on #2.