[sergo] Sergo Report: Delta-43→54 New-Linter Audit + Doc-Sync Recount - 2026-07-20

[github-actions[bot]]

Executive Summary

R61 opened onto a large registry jump: cmd/linters/main.go went from 43 → 54 registered analyzers (+11 in the 12 days since R60), and the team had already closed every prior sergo issue. Cross-referencing the 11 new linters against filed-issue history, exactly two had never been audited — timenowsub and appendoneelement — so both got a deep single-file audit. appendoneelement is fully clean; timenowsub is clean at its core but yielded two distinct findings (one recurring doc-sync drift, one narrowness FN). 2 issues filed.

Tool / Registry Updates

Signal	R60	R61
Serena LSP tools	23	23 (stable)
Registered analyzers (grep -c Analyzer main.go)	43	54 (+11)
Open sergo issues (pre-run)	0	0 (team closes fast)

11 new linters since R60: appendoneelement, bytesbufferstring, bytescomparestring, ioutildeprecated, logfatallibrary, mapclearloop, mapdeletecheck, nilctxpassed, stringscountcontains, timenowsub, trimleftright. Nine already had issues filed & closed (e.g. #44484, #45379, #45903, #46130, #46341, #46526). The two remaining — timenowsub, appendoneelement — were this run's targets.

Strategy — explicit 50/50 split

Cached component (proven "new-linter audit" vein): the highest-yield recurring Sergo strategy. Adapted to the +11 delta by diffing new linters against closed-issue titles to isolate the unaudited subset, then applying the standard checklist per linter: autofix compile-correctness, RunWithSuggestedFixes parity (#43313 class), type-resolution vs syntactic package matching (#43934 class), pattern-set narrowness, and nolint/generated-file wiring.

New-exploration component: a syntactic-stdlib-match regression sweep across all 11 new linters (grep for .Name == "<stdpkg>" identifier matching that ignores shadowing/aliasing) to confirm the #43934 astutil.IsPkgSelector migration held for newly-added code — plus the recurring doc-sync recount (header/bullet count vs registry).

Run targets: ≥1 strong distinct finding; 0 duplicates (verified all sergo issues closed); 1–3 issues.

Findings & Evidence

appendoneelement — audited CLEAN (no issue)

append(s, []T{x}...) → append(s, x).

• Autofix provably compiles. For the spread to typecheck, []T{x} must be assignable to []S (S = element type of s); slice assignability requires T == S, so x is always assignable to s's element type and append(s, x) is well-typed. No named-vs-exact-type trap (contrast writebytestring writebytestring (new 43rd linter): autofix emits non-compiling io.WriteString(w, s) for named string types + missing RunWithSugg [Content truncated due to length] #44187).
• Type-resolved: gates append via types.Universe.Lookup("append") (line 57) — no shadowing FP (testdata goodShadowedAppend).
• Correctly excludes keyed literals ([]int{5:3}), multi-element literals, and type-elided nested literals ([][]int{{1}}).
• RunWithSuggestedFixes + golden; nolint + generated-file skip wired.

Finding 1 (#46707) — doc-sync drift: doc.go "53" vs 54 registered

• pkg/linters/doc.go:3 — // All 53 active analyzers: with exactly 53 // - bullets.
• cmd/linters/main.go registers 54 (timenowsub.Analyzer at line 125 has no bullet — list skips timeafterleak → timesleepnocontext → tolowerequalfold).
• Recurring class: doc-sync: pkg/linters/doc.go says "52 active analyzers" but 53 are registered — trimleftright is omitted #46527 (52→53), doc-sync: pkg/linters/doc.go says "45 active analyzers" but 51 are registered — 6 newest linters omitted #46131 (6 omitted at once), doc-sync: pkg/linters/doc.go says "44 active analyzers" but 45 are registered — the list omits stringscountcontains #45185, doc-sync: pkg/linters/doc.go says "29 active analyzers" but 30 are registered — list omits hardcodedfilepath and sprintferrdot #40436.
• Fix: add the timenowsub bullet + bump header to 54; suggested a CI guard asserting bullet_count == registered_count to retire the class.

Finding 2 (#46708) — timenowsub isSafeSinceArg narrowness (FN)

• pkg/linters/timenowsub/timenowsub.go:131-140 — isSafeSinceArg accepts only *ast.Ident / *ast.ParenExpr, rejecting side-effect-free selectors (s.start, p.startTime).
• Excluding calls is correct (evaluation-order/side-effect safety, per testdata goodCallExprArg), but a field/pkg-var read is pure and evaluation-order-invariant against the no-side-effect Now(), so time.Now().Sub(s.start) → time.Since(s.start) is a safe simplification being missed.
• Impact latent: 0 current time.Now().Sub( sites in pkg/; non-CI-enforced simplification linter. Class: pattern_set_too_narrow (cf. lenstringzero coverage gap: len(s) != 0 is flagged but the identical idiom len(s) > 0 (and len(s) < 1 / <= 0) escapes — relation [Content truncated due to length] #40581, errstringmatch coverage gap: only strings.Contains(err.Error(), ...) is flagged — HasPrefix/HasSuffix/EqualFold/Index on err.Err [Content truncated due to length] #40244).
• Fix: add case *ast.SelectorExpr: return true + testdata pair.

Sweep result: syntactic-stdlib-match regression sweep across all 11 new linters found zero holdouts — the #43934 IsPkgSelector migration held; timenowsub (ObjectOf/PkgName) and appendoneelement (Universe.Lookup) are both properly type-resolved.

Generated Tasks

1. Doc-sync fix (doc-sync: pkg/linters/doc.go says "53 active analyzers" but 54 are registered — timenowsub is omitted #46707) — add timenowsub bullet + header→54; optional CI count-guard. Effort: trivial.
2. timenowsub coverage (timenowsub coverage gap: isSafeSinceArg rejects side-effect-free selector/index args (e.g. time.Now().Sub(s.start)) — false nega [Content truncated due to length] #46708) — extend isSafeSinceArg to *ast.SelectorExpr + testdata. Effort: small.

Metrics

Metric	Value
Run	R61
Registry delta	43 → 54 (+11)
Linters deep-audited	2 (timenowsub, appendoneelement)
Findings	5
Issues filed	2 (#46707, #46708)
Duplicates skipped	0 (all sergo issues closed pre-run)
Self-scored success	9 / 10

Historical Context

The "new-linter audit" vein remains the highest-yield strategy (R58 execcommandwithoutcontext, R60 writebytestring #44187). This run confirms the team's fast close cadence — 11 linters + ~15 issues resolved in 12 days — so pre-run gh api reconciliation is essential; gh issue list --search remains broken (use gh api ...issues?labels=sergo). Doc-sync drift is now a confirmed recurring class (5th occurrence) — hence the CI-guard recommendation.

Recommendations & Next-Run Focus (R62)

• Verify doc-sync: pkg/linters/doc.go says "53 active analyzers" but 54 are registered — timenowsub is omitted #46707 / timenowsub coverage gap: isSafeSinceArg rejects side-effect-free selector/index args (e.g. time.Now().Sub(s.start)) — false nega [Content truncated due to length] #46708 landed (team is fast).
• Watch the 55th linter — grep -c Analyzer cmd/linters/main.go vs 54.
• Audit the remaining new-linter cores (trimleftright, nilctxpassed, mapdeletecheck) to confirm their already-closed issues were fixed correctly, not just closed.
• Enforce-readiness batch for the clean new linters with zero production violations.
• Land the doc-count CI guard to permanently retire the doc-sync class.

References:

• §29717907932

Generated by 🤖 Sergo - Serena Go Expert · 243.2 AIC · ⌖ 11 AIC · ⊞ 5.8K · ◷

• expires on Jul 20, 2026, 9:05 PM UTC-08:00

[github-actions[bot]]

This discussion has been marked as outdated by Sergo - Serena Go Expert.

A newer discussion is available at Discussion #46979.