docs: add KoMA-RAG worked example

[hasnain1241]

Summary

Adds a worked example against a real, non-synthetic corpus (KoMA-RAG, a
multi-agent LLM driving system built on a highway_env fork) following the
format of the existing httpx/karpathy-repos/mixed-corpus worked examples.

What's in it

• Full extraction run (graphify extract . --code-only + graphify cluster-only)
• An honest, source-verified review.md scoring 6 dimensions, following the
  existing worked-example format
• The core finding: calls made from top-level module script code (as
  opposed to inside a function/method) don't produce calls edges, which
  explains several downstream weaknesses in this corpus (fragmented
  communities, god-node ranking dominated by a vendored dependency,
  the single most important architectural connection in the repo missing
  entirely from "surprising connections")

Full details in worked/koma-rag/review.md.

[graphify-labs]

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

---

Graphify review — findings

Adds a generated GRAPH_REPORT.md for the koma-rag corpus, capturing a 996-node / 2076-edge dependency graph across 65 communities with god-node rankings, inferred surprising connections, import-cycle results, and per-community cohesion listings. Also adds a README and a reproducibility/review writeup covering corpus stats, node/edge quality scoring, and prioritized recommendations. Documentation only — no code behavior changes.

No blocking issues surfaced. 6 lower-confidence candidates did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 81 functions depend on the 81 functions this change touches.

Health — grade A; no new coupling hotspots.

Verification — 81 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

• verification_scope: 81 function(s) in the blast radius were not formally verified this run