Add comprehensive documentation guides and improve existing docs

[sunway513]

Summary

• Add MORI-EP User Guide (docs/MORI-EP-GUIDE.md): kernel types, full Python API reference for config (17 fields) and operator (all methods), usage examples, launch config modes, standard MoE compatibility, benchmarking, framework integration, build options
• Add MORI Shmem Guide (docs/MORI-SHMEM-GUIDE.md): symmetric memory concepts, PE model, two initialization methods, query APIs, memory management, P2P address translation, synchronization
• Improve MORI-IO-INTRO: Add Python API quick reference with class/method/enum tables, code example, environment config, source files; fix StatusCode enum values, session API usage, XGMI backend status
• Improve MORI-IO-BENCHMARK: Add ToC, format arguments as table, improve readability
• Improve MORI-EP-BENCHMARK: Add ToC, fix typos, improve formatting
• Improve PROFILER: Add ToC, fix warps-per-rank constant (512→4096), fix typos
• Update README.md with Documentation table, restructured Features, Quick Start example, Framework Integration table

Fact-Checked

All documentation verified against source code:

• Every API function name, parameter, and default value
• All enum values and import paths (StatusCode: 8 values verified against pybind)
• All file paths in Source Files tables
• Build options against CMakeLists.txt
• Profiler constants against constants.hpp
• Environment variable names against source

Test plan

• Verify all guide links from README resolve correctly
• Verify code examples match actual API usage patterns

[Copilot]

Pull request overview

This PR expands and refreshes MORI’s documentation set by adding new end-to-end guides (EP + Shmem) and improving existing docs/README for clearer onboarding, API discovery, and benchmarking/profiling usage.

Changes:

• Added new comprehensive guides for MORI-EP and MORI Shmem, including API references and usage examples.
• Improved MORI-IO intro/benchmark docs with quick-reference tables, ToCs, and formatting/clarity fixes.
• Updated README with a documentation index, reorganized feature presentation, and a Quick Start snippet.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
docs/PROFILER.md	Adds ToC and corrects profiler constant text; improves wording around device-side timing.
docs/MORI-SHMEM-GUIDE.md	Introduces a new Shmem guide covering concepts, init methods, APIs, env vars, and source references.
docs/MORI-IO-INTRO.md	Adds ToC, Python API quick reference tables, and an end-to-end example plus config/source file pointers.
docs/MORI-IO-BENCHMARK.md	Adds ToC, converts arguments to a table, and restructures results sections for readability.
docs/MORI-EP-GUIDE.md	Adds a new EP user guide with kernel types, config reference, operator API coverage, examples, and integration notes.
docs/MORI-EP-BENCHMARK.md	Adds ToC and refines inter-node benchmark instructions and NIC selection section.
README.md	Adds documentation index/table, refreshes Features, and introduces a MORI-EP Quick Start example.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[TianDi101]

Thanks @sunway513 for the great contribution! We will verify the doc and merge it.

[jhchouuu]

PR #181 has already covered this PR and been merged. I will now close this PR.