[eslint-refiner] ESLint Refiner daily report - 2026-09-09 #59648
Closed
Replies: 1 comment
|
This discussion has been marked as outdated by ESLint Refiner. A newer discussion is available at Discussion #59893. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Overview
This run surfaced that repo-memory tracking for the ESLint Refiner had gone stale: it last recorded state on 2026-07-08 with a tracked count of 12 rules, but
eslint-factory/src/index.tsnow registers 61 rules. A significant backlog of openeslint-factory-labeled issues already exists that this memory had never recorded. Memory has been corrected this run; going forward dedup uses the live GitHub issue list as the authoritative source rather than repo-memory's issue log.Key metrics
eslint-factory-labeled issues found on GitHub (pre-existing, not filed by tracked runs): 10Confirmed fixes
Both defects found in
require-fs-sync-try-catchon 2026-07-08 (destructured/aliasedfsbinding false negative, and catch-lesstry/finallytreated as protective) are now fixed in the sharedtry-catch-rule-utils.tsresolver, which benefits all 5 sibling*-sync-try-catchrules. The remaining known gap in that resolver family — a boundfs/child_processvariable getting reassigned after initialization — is already tracked in #59377 and #57624; not refiled.Issues filed this run
require-http-response-error-listener(the only rule at"error"severity) can skip its check entirely when the response handler is wired viareq.on("response", cb)instead of an inline callback, or when the error listener uses.addListener()instead of.on()/.once().no-duplicate-constant-valuesapplies a size-3 minimum-duplicate-group-size guard to numbers/booleans/null (small value spaces) but a flat size-2 to all strings, including low-entropy short strings like""or single-character delimiters, which share the same small-value-space risk the rule already accounts for elsewhere.Both are latent/ungrounded findings (no live false positive/negative currently fires in
actions/setup/js/**), filed following the precedent set by #59377 for proactively documented soundness gaps.Rules investigated and ruled out this run
require-getexecoutput-exitcode-check: extensively investigated across 10+ liveignoreReturnCodecall sites (git_auth_helpers.cjs,create_labels.cjs,check_workflow_recompile_needed.cjs,push_to_pull_request_branch.cjs). All resolved correctly as true positives, true negatives, or intentionally conservative skips (e.g. the arrow-concise-body return-forwarding case inpush_to_pull_request_branch.cjs:889-898). No grounded or compelling latent defect found.require-fs-sync-try-catch: re-verified only; both prior findings confirmed fixed (see above).Pre-existing open eslint-factory issues (snapshot, not filed by tracked memory)
@actions/exec's non #58738 / prefer-actions-exec-over-child-process: preferActionsExecSyncContext message overstates the migration cost as a caller-chain cas #58169 / prefer-actions-exec-over-child-process: false negative on synchronous spawnSync used for output capture #58168 — prefer-actions-exec-over-child-process: shell semantics, migration-cost messaging, spawnSync false negativeNext actions
gh apiquery-string form) as the dedup source of truth, not repo-memory's issue log.Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.
All reactions