feat(inspect): parallel discovery and configurable timeouts

[CodeMonkeyCybersecurity]

Context

`DiscoverDocker()` runs 5 discovery phases sequentially (containers, images, networks, volumes, compose files). Each has a hardcoded 30s timeout per command.

Problem

• Sequential discovery is slow: on a host with many containers + images + networks, total time can exceed 2 minutes
• 30s timeout is hardcoded (violates zero-hardcoded-values rule)
• Network inspect per-network is also N+1 (same pattern fixed for containers in fix(inspect): harden Docker discovery with batched inspect, fix find, add tests #103)

Proposal

1. Run independent discovery phases in parallel using `errgroup`
2. Add configurable timeout via `InspectorConfig` struct
3. Batch network inspect (same pattern as container batching in fix(inspect): harden Docker discovery with batched inspect, fix find, add tests #103)
4. Batch volume inspect

Expected Impact

• 3-5x faster Docker discovery on hosts with many resources
• Configurable timeouts for CI vs interactive use