docs: add esbuild benchmark results and pilot test

[jafreck]

Summary

Add benchmark results and pilot test for Lore vs baseline evaluation on the esbuild codebase.

Changes

• docs/benchmark-results-esbuild.md — Full benchmark report: aggregate metrics, per-task detail, tool usage breakdown, ground-truth corrections, and takeaways
• tests/benchmark/pilot-esbuild.test.ts — 12 benchmark tasks targeting esbuild (callers, callees, blast radius, implementations, test mapping, complexity, dependency graph, cross-file consumers, circular deps, most-called functions, composite queries, deletion impact)

Key Results (12 tasks, 1 iteration each)

Metric	Control	Lore	Delta
First-pass accuracy	25.0%	66.7%	+41.7pp
Mean tokens	7,938	7,072	-10.9%
Mean tool calls	25.8	22.3	-14%
Lore tool usage	0%	83%	—

Note: 8 of 12 expected answers were fo> Note: 8 of 12 expected answers were fo> Note: 8 of 12 expected answers were fo> Note: 8 of 12 expected answers wert outputs and timing/token metrics remain valid.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.78%. Comparing base (eaa10ac) to head (557685d).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #199   +/-   ##
=======================================
  Coverage   87.78%   87.78%           
=======================================
  Files          76       76           
  Lines        8610     8610           
  Branches     2708     2708           
=======================================
  Hits         7558     7558           
  Misses       1052     1052           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.