LearnNewsExamplesServices
Frontmatter
id14165
titleSelf-heal soak test: N corruption+heal cycles (accelerated clock) — prove convergence, no heal-loop, no resource leak ("weeks in miniature")
stateClosed
labels
enhancementaitestingarchitecture
assigneesneo-opus-ada
createdAtJun 27, 2026, 1:20 AM
updatedAt3:22 PM
githubUrlhttps://github.com/neomjs/neo/issues/14165
authorneo-opus-vega
commentsCount2
parentIssue14039
subIssues[]
subIssuesCompleted0
subIssuesTotal0
contentTrust
projected
quarantined0
signals[]
blockedBy[]
blocking[]
closedAtJun 27, 2026, 11:01 PM

Self-heal soak test: N corruption+heal cycles (accelerated clock) — prove convergence, no heal-loop, no resource leak ("weeks in miniature")

Closed v13.1.0/archive-v13-1-0-chunk-7 enhancementaitestingarchitecture
neo-opus-vega
neo-opus-vega commented on Jun 27, 2026, 1:20 AM

Problem

#14046 / #14080 prove the immune system for a single injected corruption; #14088 adds per-class injectors. Neither proves the system survives sustained operation — the actual "runs autonomously for weeks (~95%)" bar. A heal-loop, a partial-heal residue accumulation, or a resource leak only manifests over many cycles.

Fix shape

  • A soak harness: drive N corruption → detect → classify → heal cycles on an accelerated clock against isolated stores.
  • Assert: convergence (each cycle resolves or contains), no infinite heal-loop (anti-thrash #14142 holds under load), no monotonic resource growth (WAL / snapshot / ledger bounded).

Acceptance Criteria

  • Soak harness runs ≥N mixed-class corruption+heal cycles end-to-end.
  • Asserts convergence + bounded resource use + no heal-loop across the run.
  • Reuses the #14088 / #14046 injector patterns (no parallel harness).

Scope

v13.1 — the "weeks in miniature" proof. Parent #14039. Builds on #14088 (per-class injectors) + #14142 (anti-thrash).

Live latest-open sweep: scanned 25-latest + #14046 / #14080 / #14088; those are single-shot / per-class, no soak / long-run. No duplicate.

Authored by Vega (Claude Opus 4.8, Claude Code). Origin Session ID: 1bb8a27b-ae0d-4668-a9a2-acbbe2387512