doc: list the agent backend and Azure Linux chapter in the source tree

[Pedro Henrique Penna (ppenna)]

Summary

doc/project-structure.md predates two additions to the tracked tree: the opt-in native-agent backend of the aci_edge_sandboxes crate (added in 2113c65) and the Azure Linux guest chapter in doc/guests/. This PR brings the source-tree guide back in line with git ls-files. It changes documentation only.

Base: origin/dev at 524a62699fb1607fd299f9995b846508cd49d656. No design chapter changed, so the OpenVMM pin (e5bf79776bb32f8fa29a69e7b3c9bf668477a13a) was not consulted for claims.

Changes

Statement	Before	After	Evidence
doc/guests/ entry	Guest userland chapters (Ubuntu)	(Ubuntu and Azure Linux)	git ls-files doc/guests lists azurelinux-guest.md and ubuntu-guest.md
aci_edge_sandboxes/src/ subtree	openvmm/ and bin/ only	adds agent/, an opt-in backend over a separately supplied native agent library	git ls-files aci_edge_sandboxes/src/agent; aci_edge_sandboxes/src/agent/mod.rs module docs; the agent feature in aci_edge_sandboxes/Cargo.toml
examples/ entry	Runnable lifecycle example	lifecycle examples for the OpenVMM and agent backends	[[example]] entries lifecycle (feature openvmm) and agent_lifecycle (feature agent) in aci_edge_sandboxes/Cargo.toml
tests/ entry	Mock, fake-OpenVMM, and real-hypervisor tests	adds validation and agent-library tests	aci_edge_sandboxes/tests/validation.rs and tests/agent_guest.rs module docs
Repository-layout table and aci_edge_sandboxes/ section	one OpenVMM backend	also names the opt-in agent feature's backend, a thin client of a native library that owns the sandbox lifecycle	aci_edge_sandboxes/Cargo.toml feature comment; aci_edge_sandboxes/README.md Architecture section

Follow-ups

Left out of scope; not changed here:

• scripts/setup/README.md is documentation outside doc/. Moving it needs link updates in doc/ci.md and doc/setup.md, which open PRs filesystem: map host paths from any volume, each with its own access #459, Adopt bind-once microVM image slots #406, and test-microvm: pause and resume a microVM through OpenVMM state control #405 touch, so the relocation should wait until they land.
• unittest_results.txt is a tracked unit-test log at the repository root, present since the initial commit. It looks like a stray artifact. This PR leaves it undocumented; removing it is a code-owner decision.
• python scripts/nvx.py --help still describes init as "initialize the private submodule", but the OpenVMM submodule is public (see doc/contribute.md#access). That is help text in code, so it is reported here rather than changed.
• doc/openvmm-upstream-roadmap.md records fork and upstream SHAs as of 2026-10-01, and OpenVMM pins have moved several times since. Refreshing it needs a full fork-history analysis.

Validation

• python -m unittest scripts.test_nvx_tools.CliTests scripts.test_performance.PerformanceTests: Ran 110 tests, OK (skipped=1)
• Throwaway link checker kept outside the repository: every relative link and anchor resolves in doc/project-structure.md and the files that link to it. The file has one H1 and balanced code fences, and doc/design.md links every chapter in doc/design/. Result: OK
• git diff --check: clean
• python scripts\nvx.py verify: source tree and submodule metadata are consistent
• Every changed path matches ^(doc/|.*\.md$), so CI classifies the PR as documentation-only.

[Copilot]

🟢 Approval recommended

The documentation accurately reflects the corresponding source tree, features, examples, and tests.

0 open findings

What changed in this PR

Updates the repository structure guide to reflect existing guest documentation and the opt-in native-agent backend.

Changes:

• Adds Azure Linux to the guest documentation listing.
• Documents the agent backend, example, and associated tests.
• Clarifies the crate’s dual-backend architecture.

File	Description
doc/​project-structure.md	Synchronizes the documented tree and backend overview with tracked sources.

🧠 Review effort: Balanced

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.