docs: add async SingleLatest example

[nvphungdev]

Summary

• Add examples/hello-single-latest-async with an async .source() / .tap() demo for BufferCfg::SingleLatest.
• Wire the example into the workspace, Cargo.lock, and Makefile formatting/example targets.
• Link the SingleLatest buffer table entry in the root README to the new example.

Closes #100

Validation

• cargo run -p hello-single-latest-async
• cargo fmt -p hello-single-latest-async -- --check
• cargo clippy -p hello-single-latest-async -- -D warnings
• make fmt-check
• git diff --check

Not run locally

• make check / make all: this local machine does not have the thumbv7em-none-eabihf target installed, and those targets also cover the full embedded/security workspace outside this docs/example change.

[lxsaah]

Review: PR #112 — async SingleLatest example

The PR itself is correct and well-scoped — verified locally: cargo build / cargo run (stdout matches the README), clippy -D warnings, fmt, and git diff --check all clean; API usage cross-checked against aimdb-core / aimdb-tokio-adapter; workspace wiring (Cargo.toml, Makefile, Cargo.lock, README) is complete.

Two minor nits below, each with a one-click committable suggestion. Both are non-blocking in substance — flagged as request-changes so they're resolved before merge.

[nvphungdev]

Thanks for the review. I applied both suggestions: made the Mailbox link relative and added the SingleLatest stdout note for overwritten early rollout values.\n\nChecked with cargo fmt --check and cargo check --manifest-path examples/hello-single-latest-async/Cargo.toml.

[lxsaah]

Thanks a lot for this contribution @nvphungdev and for turning around the review feedback so quickly. Both items are resolved cleanly and I've re-verified locally.

Merging now, feel free to pick up another issue from the list. Happy to have you working on more.