docs: migrate all wiki pages to mkdocs

[EshwarCVS]

Summary

• Add docs/architecture.md — radix router internals, compiled DI, lazy request parsing, middleware chain, zero-copy JSON, uvloop (from wiki Architecture page)
• Add docs/benchmark-methodology.md — benchmark types, hardware specs, direct ASGI protocol, CI workflow, fairness notes (from wiki Benchmark Methodology page)
• Add docs/concepts/sub-interpreters.md — PEP 684/734 deep dive, concurrency API, SubInterpreterPool internals, decision flowchart, examples (from wiki Sub-Interpreters Guide)
• Enhance docs/migration-from-fastapi.md — add field aliases (msgspec rename) and migration checklist (from wiki Migration page)
• Update mkdocs.yml nav to include all new pages under Concepts and Reference sections

All wiki content is now canonical in the mkdocs site. The GitHub wiki is redundant and can be disabled.

Test plan

• MkDocs strict build passes (mkdocs build --strict)
• All new pages appear correctly in the nav
• Docs CI workflow deploys successfully

https://claude.ai/code/session_01SF7ikYNfez1eVEK9d473Ef

[github-actions]

Benchmark results

Ubuntu runner, Python 3.13. HTTP table uses the same httpx load against uvicorn (Python) and Fiber (Go). Direct ASGI (below) is Python-only and excludes network I/O.

🟢 Benchmark status: improvement detected.

HTTP throughput (FasterAPI vs FastAPI vs Fiber)

Endpoint	FasterAPI	FastAPI	Fiber (Go)	F / Fast
GET /health	480/s	490/s	488/s	0.98x
GET /users/{id}	485/s	534/s	525/s	0.91x
POST /users	458/s	461/s	458/s	0.99x

Direct ASGI (no HTTP; 50,000 iterations)

Endpoint	FasterAPI	FastAPI	Speedup	vs README ASGI ratio
GET /health	190,336/s	19,394/s	9.81x	+43.3%
GET /users/{id}	164,974/s	13,462/s	12.25x	+40.4%
POST /users	101,248/s	10,953/s	9.24x	+29.3%

Routing (radix vs regex, 1,500,000 lookups)

Router	Ops/s	Speedup	vs README
Radix	962,871	9.7x	+27.8%
Regex	99,125	1.0x

How to read this

• F / Fast = FasterAPI req/s ÷ FastAPI req/s on the same HTTP harness (higher is better).
• Fiber uses the Go app in benchmarks/fiber (same routes). Go is often several times faster than Python here; the important guard for regressions is check_regressions.py (ASGI + routing floors), which must pass in this workflow.
• vs README compares combined speedups to documented reference numbers (local machine); CI absolute req/s differs by hardware.