Auto-fix CI failures for PR #1341

[github-actions]

CI Auto-Fix

Original PR: #1341
Failed CI Run: Non-Main Branch CI/CD
Failed Job: dev-build-deploy — bun run lint (biome check .) exited with code 1

Root Cause

@biomejs/biome: "^2" in package.json floats with no committed lockfile, so CI resolved it to 2.5.4 (the biome.json schema still references 2.4.16). Biome 2.5.4 promotes lint/correctness/noUnsafeOptionalChaining to a recommended error, which failed the build with 2 errors (196 additional useOptionalChain warnings do not fail biome check).

Fixes Applied

File	Line	Fix	Rule
tests/unit/instrumentation-crash-handler.test.ts	134	process.report?.writeReport -> process.report.writeReport	noUnsafeOptionalChaining
tests/unit/instrumentation-crash-handler.test.ts	149	process.report?.writeReport -> process.report.writeReport	noUnsafeOptionalChaining

Why safe: process.report is typed as non-optional ProcessReport in @types/node 25.x, so the optional chaining was unnecessary. Dropping it satisfies the rule, typechecks cleanly, and preserves behavior (the 2 affected tests pass unchanged).

Not Auto-Fixable (out of scope)

Item	Reason
196 lint/complexity/useOptionalChain warnings across src/actions/**	Biome marks these "Unsafe fix" (behavior-changing semantics). They are warnings, not errors, so they do not break the build. Bulk-fixing would risk altering auth-check logic in many files and violates the minimal, no-behavior-change scope of a CI auto-fix.
biome.json schema version mismatch (2.4.16 vs 2.5.4) + deprecated recommended field	Info/warning only; not failing. A proper biome migrate + lockfile pinning is a separate, intentional decision for the maintainers.

Verification

• bun run lint exits 0 (was 1)
• bun run typecheck passes
• Affected test file: 9/9 tests pass
• No logic changes (test-only, 2 lines)

Auto-generated by Claude AI

Greptile Summary

This auto-fix PR removes two unnecessary ?. optional chains on process.report inside a unit test file to resolve noUnsafeOptionalChaining errors introduced by Biome 2.5.4's stricter linting. The change is test-only, two lines, and preserves all existing test behavior.

• Drops process.report?.writeReport → process.report.writeReport at two assertion sites (lines 134 and 149) so the lint rule is satisfied; process.report is non-optional at runtime in Node.js and typed as such in @types/node 25.x.
• Leaves the beforeEach conditional guard (if (process.report && ...)) unchanged, creating a minor inconsistency: the guard implies process.report could be absent, while the test body now accesses it unconditionally.

Confidence Score: 4/5

Safe to merge; two-line test-only change that unblocks CI without altering runtime logic.

The fix is surgical and correct — process.report is always available in Node.js and the optional chains were genuinely unsafe per the Biome rule. The only notable issue is the inconsistency left between the guarded mock setup in beforeEach and the now-unconditional property access in the assertions, which could confuse future maintainers or cause a cryptic crash if the guard condition ever becomes relevant.

The beforeEach guard at line 79 of tests/unit/instrumentation-crash-handler.test.ts should be revisited alongside the new unconditional accesses for consistency.

Important Files Changed

Filename	Overview
tests/unit/instrumentation-crash-handler.test.ts	Removed unnecessary optional chaining on process.report?.writeReport (2 sites) to satisfy Biome 2.5.4's noUnsafeOptionalChaining rule; leaves a minor inconsistency between the guarded beforeEach mock setup and the now-unconditional test assertions.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[beforeEach] --> B{process.report && writeReport is function?}
    B -- Yes --> C[vi.spyOn writeReport mockReturnValue]
    B -- No --> D[No spy set up]
    C --> E[Test body runs uncaughtException handler]
    D --> E
    E --> F[process.report.writeReport accessed directly]
    F -- process.report defined --> G[Access .mock.calls - OK]
    F -- process.report undefined --> H[TypeError - test crashes]
    G --> I[Assertions pass]

Loading

%%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%%
flowchart TD
    A[beforeEach] --> B{process.report && writeReport is function?}
    B -- Yes --> C[vi.spyOn writeReport mockReturnValue]
    B -- No --> D[No spy set up]
    C --> E[Test body runs uncaughtException handler]
    D --> E
    E --> F[process.report.writeReport accessed directly]
    F -- process.report defined --> G[Access .mock.calls - OK]
    F -- process.report undefined --> H[TypeError - test crashes]
    G --> I[Assertions pass]

Loading

Comments Outside Diff (1)

1. tests/unit/instrumentation-crash-handler.test.ts, line 79-81 (link)

   Guard in beforeEach contradicts unconditional access in test body

   beforeEach conditionally mocks process.report.writeReport only when process.report is truthy and writeReport is a function (lines 79–81). After this fix, lines 134 and 149 access process.report.writeReport unconditionally. If the beforeEach guard ever evaluates to false — process.report absent or writeReport missing — the spy is never set up, and the direct access will throw a TypeError instead of producing a failing assertion. The PR's stated justification (process.report is non-optional in @types/node 25.x) makes this theoretical, but the guard itself now reads as dead-code that conflicts with the test's own assumptions. Consider either removing the guard (replacing with a plain vi.spyOn) or restoring optional chaining to the assertions so the two access patterns are consistent.

   Prompt To Fix With AI

   This is a comment left during a code review.
   Path: tests/unit/instrumentation-crash-handler.test.ts
   Line: 79-81
   
   Comment:
   **Guard in `beforeEach` contradicts unconditional access in test body**
   
   `beforeEach` conditionally mocks `process.report.writeReport` only when `process.report` is truthy and `writeReport` is a function (lines 79–81). After this fix, lines 134 and 149 access `process.report.writeReport` unconditionally. If the `beforeEach` guard ever evaluates to false — `process.report` absent or `writeReport` missing — the spy is never set up, and the direct access will throw a `TypeError` instead of producing a failing assertion. The PR's stated justification (`process.report` is non-optional in `@types/node` 25.x) makes this theoretical, but the guard itself now reads as dead-code that conflicts with the test's own assumptions. Consider either removing the guard (replacing with a plain `vi.spyOn`) or restoring optional chaining to the assertions so the two access patterns are consistent.
   
   How can I resolve this? If you propose a fix, please make it concise.

Prompt To Fix All With AI

Fix the following 1 code review issue. Work through them one at a time, proposing concise fixes.

---

### Issue 1 of 1
tests/unit/instrumentation-crash-handler.test.ts:79-81
**Guard in `beforeEach` contradicts unconditional access in test body**

`beforeEach` conditionally mocks `process.report.writeReport` only when `process.report` is truthy and `writeReport` is a function (lines 79–81). After this fix, lines 134 and 149 access `process.report.writeReport` unconditionally. If the `beforeEach` guard ever evaluates to false — `process.report` absent or `writeReport` missing — the spy is never set up, and the direct access will throw a `TypeError` instead of producing a failing assertion. The PR's stated justification (`process.report` is non-optional in `@types/node` 25.x) makes this theoretical, but the guard itself now reads as dead-code that conflicts with the test's own assumptions. Consider either removing the guard (replacing with a plain `vi.spyOn`) or restoring optional chaining to the assertions so the two access patterns are consistent.

_{Reviews (1): Last reviewed commit: "fix(lint): resolve noUnsafeOptionalChain..." | Re-trigger Greptile}