feat: --estimate flag on issues create and issues update

[lightstrike]

Summary

• Add --estimate / -e flag to issues create and issues update commands
• Wire estimate into IssueCreateInput and IssueUpdateInput
• Update no-op guard on issues update to include --estimate
• Add [-e N] to lineark usage for both commands
• Add 3 offline tests for estimate flag visibility and error messages

Test plan

• cargo clippy --workspace -- -D warnings clean
• cargo fmt --check clean
• cargo test --workspace — 67 offline tests pass (3 new)
• Smoke test: lineark issues create "Test" --team ENG --estimate 3
• Smoke test: lineark issues update ENG-xxx --estimate 5

[flipbit03]

@lightstrike column alignment needs to be fixed in some places. usage.rs column alignment is broken. I've had this issue happen so many times in Claude Code-assisted development that I've simply created a linter for it 😁 It should be trivial to fix. Thanks!

[lightstrike]

@flipbit03 Ready for review again — rebased onto main and fixed the column alignment in usage.rs. All CI checks passing.

[flipbit03]

Hey @lightstrike — I rebased onto the latest main (which now includes PR #88), resolved the conflicts, and pushed. I also made a few improvements:

Fixes applied:

• Unique test names: The online test for estimate used a hardcoded "[test] CLI estimate flag" name — changed it to use a UUID suffix to avoid conflicts from zombie resources (same fix we applied across all tests in feat: --project filter on issues list #88)
• Proper error handling: Added output.status.success() checks with stdout/stderr in assertion messages before parsing JSON, so failures show the actual error instead of cryptic parse errors
• Added estimate to issues read output: The PR wired --estimate into create/update, but the IssueDetail type used by issues read didn't include the field — so there was no way to verify the estimate via CLI after setting it. Added it so issues read now shows "estimate": 3.0.

Smoke tests confirmed:

• lineark issues create "test" --team CAD -e 3 → creates with estimate 3
• lineark issues read CAD-xxx → shows "estimate": 3.0
• lineark issues update CAD-xxx -e 5 → read confirms "estimate": 5.0
• Both -e short flag and --estimate long flag work

Follow-up issue: While reviewing the output, we noticed priority is output as both a raw number ("priority": 0.0) and a label ("priorityLabel": "No priority"), which is redundant and confusing. Carved out #116 to clean that up separately.

All 51 online tests passing. Ready to merge!

[flipbit03]

LGTM! Let's gooooooo