perf(storage): parallelize Claude Code transcript listing

[E2ern1ty]

Summary

Listing Claude Code transcripts walked the projects directory and every transcript one await at a time, so a full scan serialized its round trips through the thread pool instead of overlapping them — and as the adapter's own comments note, the catalog is listed once per search term, so the walk is paid again on files that did not change.

The directory walk, the per-file stats, and the summary derivations are now concurrent:

• listSessions derives all summaries via Promise.all instead of one await per file. #summaryOf never throws and only touches the summary cache at distinct keys (one file per session id), so the calls are independent; results keep the original order and the final sort is stable.
• #transcriptFiles lists project directories and stats candidate files concurrently. The newest-wins dedup stays a sequential, order-independent reduce (its tie-break names the path, not the iteration order), so completion order cannot change which file wins.
• #parse takes the file size the cache check already stat'ed, instead of stat'ing the same file a second time.

Measured on 1000 seeded transcripts (20 projects, ~50 records each): warm relist (all files cached) drops from ~20 ms to ~5 ms; the first listing from ~155 ms to ~120 ms.

Also adds a test pinning the previously-untested dedup rule that the concurrent walk must preserve: one session id under two project directories must resolve to the newest copy in both list and read.

Verification

• npm --workspace @maka/storage run typecheck — clean
• npm --workspace @maka/storage run test:dist — 1215 pass / 0 fail, including the new test
• npx biome check on both changed files — clean
• Before/after benchmark (mean of 2 runs each, 1000 transcripts): warm listSessions 20–22 ms → 5–8 ms; cold 155–156 ms → 118–128 ms

AI use

Select exactly one:

• No generative tool made a substantive contribution
• Generative tooling made a substantive contribution

Tool(s) and scope: ZCode authored the change, the new test, and this description; benchmark numbers were measured locally on real runs.

Checklist

• Tests cover the change and fail without it
• Lint, format, typecheck and the affected suites pass locally

Note: this is a behavior-preserving performance change, so no test fails without it; the new test pins the newest-wins dedup invariant the concurrent walk must uphold.

Does this PR entail a change in behavior?

• Yes — described under Summary above
• No