Retry npm install for disable wasm integration test

[masih]

The Disable WASM integration shard was flaky because the test script ran a single bare npm ci before starting Hardhat. In the failed CI job, npm aborted with a transient ECONNRESET while fetching packages, so the shard failed before DisableWasmTest.js actually ran.

Retry npm ci with a short backoff and remove any partial node_modules contents between attempts. Persistent install failures still fail the job, but temporary registry/network resets no longer make the test shard flaky.

Flaked in unrelated changes.

[cursor]

PR Summary

Low Risk
Test harness script only; no application, auth, or chain logic changes.

Overview
The Disable WASM integration test script no longer runs a single npm ci before Hardhat. It now tries up to three installs, waits 5s / 10s between failures, and removes node_modules before each retry so partial installs do not carry over.

Hardhat still runs only after a successful install; a third failed attempt exits with failure unchanged from before, aside from the extra resilience against transient registry errors (e.g. ECONNRESET).

^{Reviewed by Cursor Bugbot for commit 9399910. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

The latest Buf updates on your PR. Results from workflow Buf / buf (pull_request).

Build	Format	Lint	Breaking	Updated (UTC)
✅ passed	✅ passed	✅ passed	✅ passed	Jul 1, 2026, 11:50 AM

[seidroid]

A small, correct CI hardening change that wraps npm ci in a 3-attempt retry loop with backoff to address a flaky integration shard. The shell logic correctly handles set -e, fails fast on persistent errors, and cleans partial installs between attempts.

Findings: 0 blocking | 3 non-blocking | 0 posted inline

Blockers

• None at the file/PR level.

Non-blocking

• Cursor's second-opinion review file (cursor-review.md) was empty/produced no output; only Codex provided a pass (no material findings). The repo REVIEW_GUIDELINES.md was also empty, so no repo-specific standards were applied.
• Optional: the same flaky-install pattern likely affects other integration scripts that run a bare npm ci; consider extracting a shared retry helper rather than duplicating this loop if the pattern recurs.
• Nit: rm -rf node_modules between attempts is largely redundant because npm ci already removes node_modules before installing — harmless, but worth noting it doesn't clear a corrupt npm cache, which is another common transient-failure cause (npm cache clean --force or removing ~/.npm could help if ECONNRESET corrupts a cached tarball).

[claude]

LGTM, simple CI flake mitigation with retry/backoff for npm ci.

Extended reasoning...

Overview

This PR modifies a single CI shell script, integration_test/evm_module/scripts/disable_wasm.sh, to wrap npm ci in a retry loop (up to 3 attempts) with a 5/10-second linear backoff and node_modules cleanup between attempts. The motivation is a documented flaky CI run where npm aborted with a transient ECONNRESET before the actual Hardhat test could execute.

Security risks

None. This is a CI-only shell script not used at runtime. The retry logic does not introduce any new inputs, network endpoints, or privilege changes. rm -rf node_modules operates within the script's working directory (contracts/) so the blast radius is limited.

Level of scrutiny

Low. This is a mechanical, defensive change to a CI helper script — the kind of routine flake mitigation that does not warrant human review. The control flow is straightforward, set -e semantics are preserved (the if npm ci guards the failure correctly), and the final exit 1 ensures persistent failures still surface.

Other factors

The bug hunting system found no issues. The change is minimal, well-scoped, and the PR description clearly documents the symptom and remediation. No production code paths are touched.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 58.00%. Comparing base (42d7e20) to head (9399910).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3674      +/-   ##
==========================================
- Coverage   59.28%   58.00%   -1.28%     
==========================================
  Files        2272     2179      -93     
  Lines      188193   177388   -10805     
==========================================
- Hits       111567   102902    -8665     
+ Misses      66570    65352    -1218     
+ Partials    10056     9134     -922     

Flag	Coverage Δ
sei-db	70.41% <ø> (ø)
sei-db-state-db	?

Flags with carried forward coverage won't be shown. Click here to find out more.
see 151 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[seidroid]

The PR wraps npm ci in a 3-attempt retry loop with linear backoff and node_modules cleanup to de-flake the Disable WASM integration shard. The logic is correct — npm ci runs inside an if so set -e doesn't abort early, the 3rd failure still exits non-zero, and the Hardhat invocation is unchanged.

Findings: 0 blocking | 2 non-blocking | 0 posted inline

Blockers

• None at the file/PR level.

Non-blocking

• Cursor's review (cursor-review.md) produced no output (empty file); this synthesis is based on my own analysis plus Codex, which also reported no material findings.
• Optional robustness nit: since the retry targets transient registry/network errors, you could add npm cache clean --force alongside rm -rf node_modules between attempts to also recover from a corrupted npm cache. Not required for correctness.

[claude]

LGTM — straightforward retry loop on npm ci to deflake the Disable WASM integration shard.

Extended reasoning...

Overview

This PR modifies a single integration test shell script, integration_test/evm_module/scripts/disable_wasm.sh, wrapping the existing npm ci invocation in a 3-attempt retry loop with backoff (5s, 10s) and rm -rf node_modules between attempts. If all three attempts fail, the script exits with status 1, preserving the prior failure semantics. set -e remains in effect, and npx hardhat test still only runs after a successful install.

Security risks

None. This is a CI-only shell script that runs during integration testing; it does not touch any application, chain, auth, crypto, or consensus code. The retry logic uses static integers and does not introduce any injection surface.

Level of scrutiny

Low. Test infrastructure flake fixes of this shape (retry a network-bound step, clean partial artifacts between attempts) are mechanical and well-understood. The scripted if npm ci; then break; fi pattern is compatible with set -e, and the third-attempt exit is explicit.

Other factors

No bugs were flagged by the bug hunting system. The Cursor Bugbot summary and CI checks (Buf) pass. The change is small, self-contained, and reverts cleanly if it doesn't help.