test(cli): cover functions offline serve e2e with effect lint (CLI-2592)

[7ttp]

TL;DR

brings the offline functions serve e2e under the effect lint and covers all of shared/functions

whats introduced?

• the 21 per file allow list entries collapse into one !apps/cli/src/shared/functions/**, which also covers the untouched serve.main.ts, serve.errors.ts and serve-main-deps.ts
• the e2e drives docker through ChildProcessSpawner and HTTP through HttpClient with the same request bytes, and polls with Schedule against the same deadlines
• containers, the network and temp dirs are cleaned up through the test scope, also when a run is aborted, and the network goes even when container removal fails
• all three tests are kept, a hung request still reports the container diagnostics, and a timed out docker logs keeps its partial output

ref

• closes: CLI-2592

[github-actions]

🤖 AI Review

Both independent reviews completed; Claude reported two findings and Codex reported none. Code inspection confirms two minor issues: timed-out log collection loses partial output, and an effect-level failure during container removal skips network cleanup. Cleanup timeouts were already absent before this PR. Finding locations were corrected to match the checked-out code. Verification was static because dependencies are unavailable.

Findings

Severity	Location	Category	Sources	Claim
🟡 MINOR	apps/cli/src/shared/functions/serve-main-offline.e2e.test.ts:186	diagnostics	claude	When docker logs times out, containerLogs discards any stdout/stderr already received and returns only a failure marker, reducing failure diagnostics.
🟡 MINOR	apps/cli/src/shared/functions/serve-main-offline.e2e.test.ts:364	resource-cleanup	claude	An effect-level failure during docker rm skips the subsequent network-removal attempt, weakening best-effort cleanup.

Stats

Claude findings: 2 · Codex findings: 0 · Confirmed: 2 · Refuted: 0 · Uncertain: 0

---

Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: auto · Workflow run

This review runs once per PR. A maintainer can request another with a /ai-review comment.