Post-merge fixes for bucket_list (PR #285)

[drernie]

Overview

Copilot review of PR #285 (submitted after merge) identified several issues in the bucket_list implementation.

Critical Fixes

1. Validate bucket name is present

File: src/quilt_mcp/tools/buckets.py

Currently defaults to empty string: bucket_data.get("name", "") - this silently creates invalid BucketConfig entries where name is required.

Fix: Validate name is present/non-empty, return BucketListError if missing.

2. Handle GraphQL errors explicitly

File: src/quilt_mcp/tools/buckets.py

Currently ignores result.get("errors") and treats error responses as successful empty lists.

Fix: Check for "errors" in result after execute_graphql_query(), return BucketListError with server messages.

Important Fixes

3. Fix camelCase inconsistency

File: src/quilt_mcp/tools/responses.py

BucketConfig uses camelCase (iconUrl, relevanceScore) while all other response models use snake_case.

Fix: Rename to icon_url / relevance_score (use Pydantic alias= if needed for GraphQL compatibility).

4. Update test expectations for GraphQL errors

Files:

• tests/unit/tools/test_buckets.py:347
• tests/func/test_bucket_list_func.py:219

Tests expect success=True when GraphQL returns errors. Should expect BucketListError instead.

Medium Priority

5. Fix boolean precedence bug in test-runner

File: scripts/test-runner.py:558

is_actual_error logic: ... or "FAILED" in line_clean has wrong precedence, captures FAILED lines even when excluded.

6. Relax E2E optional field assertions

File: tests/e2e/tools/test_bucket_list_e2e.py:98

Enforces non-null for optional fields (title, relevanceScore, browsable) - may cause flaky failures.

Low Priority

7. Include security tests in CI

File: tests/security/test_bucket_list_auth.py

Security tests not run by make test-ci (only runs unit/func/e2e).

---

Priority: Critical fixes (#1, #2) should be addressed immediately. Important fixes (#3, #4) should be included in the same PR.