✨ Split SDK helper contracts

[Robdel12]

Why

The SDK and browser client are the public entrypoints people build around, so they need to be the most boring part of the CLI: stable config handling, predictable helper behavior, and type definitions that match the actual runtime. Before this split, a lot of that logic lived inline or behind implicit behavior, which made later command and server cleanup harder to review safely.

This PR establishes the small shared helper contracts first. That gives the rest of the stack a clean base for URL normalization, screenshot option normalization, timeout behavior, wildcard matching, and SDK service creation without pulling unrelated command/server changes into the same review.

What Changed

• Tightened the SDK/client surface and matching type definitions.
• Added focused helpers for API URL derivation, async timeouts, fetch timeouts, wildcard matching, and screenshot property normalization.
• Added regression coverage for SDK config updates and helper behavior.
• Updated type tests so published typings track the new public surface.

Verification

• npm run build
• npm run test:types
• node --test tests/utils/api-url.test.js tests/utils/async-utils.test.js tests/utils/fetch-utils.test.js tests/utils/patterns.test.js tests/utils/screenshot-options.test.js tests/sdk/client.test.js tests/sdk/index.test.js tests/api/endpoints.test.js

Stack

1/10 in the CLI audit stack.

Base: main
Head: rd/cli-audit-sdk-helpers

[vizzly-testing]

Vizzly - Visual Test Results

CLI Reporter - 1 change needs review

Status	Count
Passed	18
Changed	1
Auto-approved	18

Changes needing review (1)

fullscreen-viewer · Firefox · 375×667 · 78.5% diff

Review changes

CLI TUI - Approved

5 comparisons, no changes detected.

View build

---

_{rd/cli-audit-sdk-helpers · 8b9bcbf0}