Add retry logic for transient GitHub API errors in live integration test

[Copilot]

The js-integration-live-api workflow was failing on transient GitHub API errors: "No server is currently available to service your request". These errors should be retried automatically.

Changes

• Extended transient error detection (error_recovery.cjs)

  • Added "no server is currently available" pattern to isTransientError()

• Wrapped API calls with retry logic (frontmatter_hash_github_api.test.cjs)

  • Created createRetryableFileReader() wrapper with 3 retries, 1s initial delay, exponential backoff
  • Applied to live GitHub API file reader

• Added test coverage (error_recovery.test.cjs)

  • Verified "no server" errors are identified as transient

// Before: Direct API call
const fileReader = createGitHubFileReader(octokit, owner, repo, ref);

// After: Wrapped with retry logic
const baseFileReader = createGitHubFileReader(octokit, owner, repo, ref);
const fileReader = createRetryableFileReader(baseFileReader);

The test now handles temporary GitHub API unavailability without masking real failures.

Original prompt

Fix the failing GitHub Actions workflow js-integration-live-api
Analyze the workflow logs, identify the root cause of the failure, and implement a fix.
Job ID: 63124286802
Job URL: https://github.com/github/gh-aw/actions/runs/21870586097/job/63124286802

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[Copilot]

Pull request overview

Improves resilience of the js-integration-live-api workflow/tests by expanding transient GitHub API error detection and adding retry/backoff around live API file reads.

Changes:

• Extended isTransientError() to treat “no server is currently available” GitHub API failures as transient.
• Wrapped the live GitHub API file reader in frontmatter_hash_github_api.test.cjs with withRetry() using an exponential backoff configuration.
• Added unit coverage confirming the new transient-error pattern is detected.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
actions/setup/js/frontmatter_hash_github_api.test.cjs	Adds retry wrapper + config for live API reads; injects global.core for retry logging.
actions/setup/js/error_recovery.test.cjs	Adds assertions that “no server is currently available” is treated as transient.
actions/setup/js/error_recovery.cjs	Adds the new transient error substring pattern to isTransientError().

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

LIVE_API_RETRY_CONFIG.initialDelayMs is likely not the actual delay before the first retry with the current withRetry implementation: withRetry multiplies delay at the end of the first failure, so the first retry sleeps initialDelayMs * backoffMultiplier (default 2). If the intent is a 1s first backoff, either pass backoffMultiplier: 1 / adjust initialDelayMs, or change withRetry to increase the delay after sleeping so initialDelayMs is honored for the first retry.

Suggested change

	initialDelayMs: 1000,
	initialDelayMs: 1000,
	backoffMultiplier: 1,

[Copilot]

This test sets global.core in beforeAll but never restores the previous value. Since Vitest may execute multiple test files in the same worker, this can leak logging mocks into other suites. Please save the prior global.core value and restore it in an afterAll/afterEach (or delete global.core if it was unset).