| /plans/token-efficient-retrieval.md |
0.90 |
Plan: Token-efficient retrieval |
plan, token-efficiency, retrieval, lookup, match-body, section-first, ranking, hubs, hub-gate, curation, context, budget, lean-envelope, dedup, promote, benchmark, tokens-per-answer, ecosystem-optimization, focus, status:active |
| /plans/graph-focus.md |
0.85 |
Graph focus: correct, current, useful context |
plan, graph, performance, relevance, usability, context, retrieval, benchmark, freshness, provenance, tenant-isolation, status:done |
| /adr/0012-lookup-body-match.md |
0.85 |
ADR 0012: Body match is an optional mode of LOOKUP |
adr, decisions, lookup, body-match, match-body, section-index, full-text, search, snippets, anchors, spec, protocol, catalog, mdoutline, token-efficiency, retrieval, ranking |
| /plans/graph-focus/retrieval-recovery.md |
0.80 |
Retrieval recovery plan |
graph, retrieval, benchmark-calibration, failure-audit, recovery, roadmap |
| /benchmarks.md |
0.80 |
Benchmarks |
benchmarks, tokens, cost-per-query, retrieval, evaluation, graphifymd, ckg, model-tokens |
| /plans/token-efficient-retrieval/design-benchmark.md |
0.80 |
Token-efficient retrieval: benchmark |
plan, benchmark, retrieval, tokens-per-answer, baseline, hit-rate, token-efficient-retrieval |
| /plans/graph-focus/first-hit-lookup.md |
0.75 |
First-hit lookup expansion |
plan, lookup, first-hit, expansion, calls-per-question, retrieval, benchmarks, mcp |
| /plans/graph-focus/evaluation.md |
0.75 |
Graph focus evaluation |
graph, benchmark, retrieval, evaluation, relevance, tokens, performance, usability, status:proposed, end-to-end, provider-usage, real-corpus, soul, section-first, reader-policy |
| /research/graphifymd-comparison.md |
0.75 |
GraphifyMD comparison |
research, graphifymd, compressed-knowledge-graph, knowledge-graph, context, retrieval, token-efficiency, evidence-subgraph, multi-hop, benchmark, provenance, typed-relations, architecture |
| /plans/token-efficient-retrieval/hub-gate-2026-09-08.md |
0.70 |
Hub gate delta 2026-09-08 |
benchmark, retrieval, delta, hub-gate, hub-size, style-gate, index-split, token-efficiency, tokens-per-answer, lookup, body-match, d06 |