Roka Docs
Concepts
Benchmark
MVP study: the shipped MCP prune_file tool on the LogHub corpus. 15 real production log sources, one 4,000-character budget, no per-file tuning.
| Metric | Result |
|---|---|
| Avg. lines removed | 92.2% |
| Avg. compression | ~69× (chars in → 4K out) |
| Sources with real errors still surfaced | 11 / 11 |
What this measures
Retention under budget — did the important line survive? It does not measure downstream agent task success, wall-clock RAG savings, or ranking quality vs embedding baselines.
Full methodology, Thunderbird before/after, and explicit non-claims: research.html. Per-dataset table: testing.html.