chore(cli): bring /shared under the effect skill

[7ttp]

TL;DR

most of /shared already followed the effect skill after its effect lint coverage in CLI-2458,
this picks up the few things that were left

whats introduced?

• functions deploy retries and rate limit waits move from hand written loops to Effect.retry and Effect.repeat with a Schedule, keeping the same attempts, delays and messages
• serve outcomes become Data.TaggedEnum values matched with $match, and the shutdown grace uses Effect.timeoutOrElse
• Predicate.isTagged and exhaustive matching replace _tag checks, so the trace sanitizer now handles every decision explicitly
• casts and non null assertions give way to narrowing, and consent encodes through Effect
• the feedback client and project link state use Effect.fn
• tests run on it.effect wherever the clock does not matter, and the malformed ref services test now asserts no request is sent

notes

• no user facing behaviour changes at all
• the four new spans only appear in trace output and the SUPABASE_DEBUG=1 span console

ref

• extends: chore(cli): bring the remaining commands under the effect skill (CLI-2693) #7122

[github-actions]

🤖 AI Review

Reconciled both supplied reviews and confirmed one minor test-coverage gap. No runtime defect was established. Effect dependency source is unavailable, so the schedule metadata assumptions remain unverified; tests were not run.

Findings

Severity	Location	Category	Sources	Claim
🟡 MINOR	apps/cli/src/shared/functions/deploy.ts:1780	test-coverage	claude	The rewritten 429 schedule lacks regression coverage for retry exhaustion, fallback backoff, and the displayed retry duration. Its behavior depends on schedule attempt numbering and whether tap receives the modified duration, but the existing test cannot detect mistakes in either assumption.

Stats

Claude findings: 1 · Codex findings: 0 · Confirmed: 1 · 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.