[eslint-refiner] ESLint Refiner — 2026-08-11: 2 issues filed on require-spawn-error-listener; memory-gap reconciled

[github-actions[bot]]

Overview

Reviewed the two ESLint factory rules that had never been reviewed before: require-spawn-error-listener (shipped 2026-08-09) and require-mkdtempsync-try-catch (shipped 2026-08-10). Filed 2 issues on the former; the latter is clean. Also discovered and reconciled a ~1-month gap in this workflow's persisted repo-memory.

Key issues filed

• DI-fallback binding blind spot — isSpawnBinding only recognizes spawn via a direct import or require() destructure, so const spawnImpl = options.spawnImpl ?? spawn; is invisible to the rule. Live at copilot_sdk_sidecar.cjs:128-131 (currently safe, but unprotected against regression). Same defect shape as the prior CORE_ALIASES DI-parameter gap (eslint-factory: core-method-resolve.ts CORE_ALIASES gap makes 8 sibling rules blind to DI-style core parameter names #49489).
• Path-insensitive listener check — hasErrorListener accepts a listener attached anywhere in scope, including inside a conditional branch that doesn't cover every path from the spawn() call. This is the third occurrence of this exact "whole-block/whole-callback branch-order false negative" defect class in this factory (previously eslint-factory: no-unsafe-promise-catch-error-property guard is whole-callback, causing branch-order false negatives #42915, eslint-factory: no-core-error-then-process-exit only checks adjacent statements — misses intervening statements (FN) #46993, eslint-factory: require-return-after-core-setfailed misses bare switch-case fall-through (FN) and flags trailing function declar [Content truncated due to length] #46992). No live occurrence found; filed proactively given the repeat pattern.

require-mkdtempsync-try-catch got a clean review — no issue filed. It's a thin wrapper over the shared try-catch-rule-utils helpers already hardened during the 2026-07-08 review of require-fs-sync-try-catch, and its own 11-case test suite already covers the tricky shapes (destructured bindings, ESM namespace import, computed access, catch-less try/finally).

Memory-gap reconciliation

This workflow's repo-memory (state.json/history.jsonl) was last updated 2026-07-08, recording 12 reviewed rules. The actual rule inventory has grown to 44 rules, and GitHub issue history (gh api search/issues?q=repo:github/gh-aw+label:eslint-factory) shows the workflow kept running and filing issues daily through 2026-08-10 — the memory push itself was silently failing or landing elsewhere for about a month, without affecting the underlying mission. Cross-referencing that issue history against eslint-factory/src/index.ts showed 42 of 44 rules already have at least one filed issue; only the two rules reviewed today had zero. state.json has been rewritten to reflect the accurate inventory and to record this reconciliation so future runs trust the GitHub issue history over stale local state.

Full grounding detail

require-spawn-error-listener — live call sites checked:

• process_runner.cjs:92 — true negative, unconditional child.on("error", ...) at line 187.
• apply_samples.cjs:627 — true positive (real app bug), no listener anywhere in file.
• start_mcp_gateway.cjs:583 — true positive (real app bug), detached+unref() child with no listener anywhere in file.
• copilot_sdk_sidecar.cjs:131 — missed entirely due to the DI-fallback binding gap (issue 1 above); has a listener at line 159 but the rule never evaluates the call site.

require-mkdtempsync-try-catch — live call sites checked:
artifact_client.cjs:371,419, generate_git_bundle.cjs:246,251, restore_aic_usage_cache_fallback.cjs:166, fuzz_template_substitution_harness.cjs:65, check_daily_aic_workflow_guardrail.cjs:280. Traced ancestor try-blocks by hand for 3 of 5; all confirmed or strongly indicated as protected. No novel gap found — any remaining edge case matches a known, already-accepted scope limit shared across ~10 sibling try-catch rules (rule doesn't cross function-call boundaries), not a new defect.

Other open issues: #50496 (require-nan-check-after-env-numeric-parse, ternary-wrapped env parse escapes detection) remains open — flagging as a candidate to nudge if it stays open much longer. Every other eslint-factory-labeled issue is closed.

Next

• Watch for a fix on require-spawn-error-listener's two filed issues; re-verify copilot_sdk_sidecar.cjs:131 gets covered once the DI-fallback fix lands.
• No new unreviewed rule expected tomorrow unless a 45th rule ships overnight — next run will re-check index.ts first per usual cadence.
• repo-memory now reflects the true 44-rule state; future runs should still cross-check GitHub issue history as the reconciliation showed local memory can silently drift.

Generated by 🤖 ESLint Refiner · agent · 262.1 AIC · ⌖ 28.7 AIC · ⊞ 4.7K · ◷

• expires on Aug 11, 2026, 9:48 PM UTC-08:00

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-08-12T05:48:47.250Z.

Closed by Workflow