test: increase command layer test coverage from 28% to 80%+

[iamfj]

Problem

The commands layer has 28.18% statement coverage. 9 of 11 command files have 0% coverage:

Command File	Coverage	Risk
issues.ts	50.53%	High — complex flag validation untested
auth.ts	81.81%	Low
comments.ts	0%	Medium
cycles.ts	0%	Medium
documents.ts	0%	Medium — most logic of any read-only command
files.ts	0%	Medium
labels.ts	0%	Low — simple pass-through
milestones.ts	0%	Medium
projects.ts	0%	Low — simple pass-through
teams.ts	0%	Low — simple pass-through
users.ts	0%	Low — simple pass-through

The highest-risk untested code is the issues update flag validation:

• Mutual exclusivity: --parent-ticket vs --clear-parent-ticket
• --label-mode requires --labels
• --clear-labels conflicts with --labels and --label-mode
• Relation flag mutual exclusivity (5 flags, only 1 allowed)

Approach

Command tests should mock createContext() to return mock gql/sdk clients, then verify:

1. Flag validation — invalid combinations throw before any API call
2. Resolver calls — correct identifiers passed to resolvers
3. Service calls — correct UUIDs and inputs passed to services
4. Output — outputSuccess() called with service result

Test template

vi.mock("../../src/common/context.js", () => ({
  createContext: vi.fn(() => ({
    gql: { request: vi.fn() },
    sdk: { sdk: { teams: vi.fn(), issues: vi.fn() } },
  })),
}));

Priority order

1. issues update — highest complexity, highest risk
2. issues create — many optional flags with dependencies
3. documents — has --issue attachment logic, URL extraction
4. cycles — --window flag logic
5. milestones — project-scoped resolution
6. Remaining simple commands

Acceptance criteria

• issues update flag validation: all mutual exclusivity checks tested
• issues create required flag (--team) tested
• Every command file has ≥1 happy-path + ≥1 error-path test
• Overall command coverage ≥70%
• npm run test:coverage reports improvement