nvx-rs: retire the golden CLI scenarios by ppenna · Pull Request #830 · microsoft/nvx

Pedro Henrique Penna (ppenna) · 2026-10-23T23:42:06Z

Closes #728: the second of its two pull requests. It builds on #825, which added the scenario tests, and is rebased on dev since #825 and #828 merged, so its four commits apply to dev directly.

Changes

The help of run (76b47ad): nvx.py run parses the options of nvx-rs run before it runs it, so it prints a help of its own, which a test of nvx.py compared with the help that cli.json recorded. tests/fixtures/run-help.txt now holds that help, byte for byte: run_prints_the_help_of_nvx_py_run checks that nvx-rs run prints it, and test_run_help_is_the_help_of_nvx_rs_run that nvx.py run prints it too, with the default memory of each guest of guests.py.
Coverage (e4bdbf0): scripts/nvx_rs_coverage.py no longer measures the family of the replay, whose filter would select no test, and doc/ci.md notes its retirement. With the egress family that nvx-rs: retire the egress policy fixture and read policies as strict JSON #828 retired, the Python semantics and launch profile families remain.
Retirement (58d5973): tests/fixtures/cli.json goes, with nvx_rs_reproduces_the_recorded_scenarios, its normalization and population helpers, and the mapping of tests/cli/scenarios.rs; the tables of cases are private again. The README's Golden scenarios section gives way to Scenario tests, which records how they replaced the transcripts, and .gitattributes and a test of the guest binaries' source identity no longer name the fixture: that test names root-tokens.json, which holds no behavior recorded from Python, so that no retirement of nvx-rs: retire the Python-parity golden tests and make Rust the reference #722 removes it. The scenario tests also expect the two messages that nvx-rs: retire the egress policy fixture and read policies as strict JSON #828 changed while it updated only the transcripts: an unknown field of an egress policy is quoted as Rust quotes it, "other", and a policy file that cannot be read is named as it was given, ./ included.
Configuration format (7b0feef): provision writes format 7 for every sandbox, with export null without shares, where it wrote format 1 for those, and start still reads format 1. The scenario tests start a hand-written configuration of format 7 without shares, and check that a start refuses one of format 7 without export and one of format 1 with it.

Differences from the issue

The issue asks to keep reading formats 1 to 6, but nvx-rs: host cold integration for image-backed sandboxes #622 had already stopped reading formats 2 to 6, which recorded the shares as requested; start reads format 1 and refuses those as before.
The README no longer promised that an older nvx.py could operate a sandbox that nvx-rs provisioned: start already refuses a configuration without the bindings that only nvx-rs records.

Gates

#825 holds the evidence, from runs on its branch, whose CLI sources this pull request leaves unchanged but for the format of config.json:

Lines and branches that only the replay reached are reached by the scenario tests, or depend on timing and random inputs: the replay reaches no line or branch of production code beyond the tests outside its family, on Linux or Windows (run 40028577794).
Mutants: the other tests of --test cli catch or time out on every mutant that the replay catches. On Linux, that covers 791 mutants of src/host/cli/**/*.rs, a superset of sandbox/{mod,lifecycle,state}.rs. On Windows, it covers 505 mutants of the files with platform-specific code.
Every scenario maps to a test, as nvx-rs: add the scenario tests that replace the CLI transcripts #825's mapping test checked until this pull request retired it.

Rebase on `dev`

The rebase onto dev at ae0ceb4 left the patches of the first and last commits unchanged. #828 conflicted with the other two:

nvx-rs: retire the egress policy fixture and read policies as strict JSON #828 retired egress-policy.json and its coverage family. The coverage script, doc/ci.md, .gitattributes, and the README's Python parity section now drop both families and fixtures, the retirement deletes the cli.json that nvx-rs: retire the egress policy fixture and read policies as strict JSON #828 updated, and the test of the guest binaries' source identity, which named egress-policy.json before the rebase, names root-tokens.json.
nvx-rs: retire the egress policy fixture and read policies as strict JSON #828 was tested on a base without nvx-rs: add the scenario tests that replace the CLI transcripts #825's scenario tests, and dev has failed two of them since it merged, on Linux and Windows (run 40153682491): egress_compile_prints_the_library_lowering expected 'other', and validation_errors_exit_1_before_any_side_effect a normalized path. The retirement commit updates both expectations, so the first two commits fail those two tests, as dev does, and the last two pass them.

Validation

After the rebase, on Windows, Rust 1.95.0:

At the head, every Windows step of check-nvx-rs: rustfmt; Clippy with -D warnings for all targets and features and for the library without default features; the tests with all features and with the default ones; the library without default features; the artifact bundling test; rustdoc with -D warnings; cargo +1.89 check --all-targets --all-features; and the runtime adapter's protocol-only and host-only configurations. Every Windows step of check-aci-edge-sandboxes, its agent bundling and Rust 1.89 check included.
At each commit, in a worktree of the commit: rustfmt, both Clippy configurations, rustdoc, the cli tests, Ruff, the Markdown link check, and test_nvx_tools.py, test_guest_binaries.py, test_nvx_rs_coverage.py, and test_markdown_links.py. The cli tests of the first two commits fail only the two tests that dev fails.
Python, at the head: Ruff lint and format, Pyright for Linux and Windows, the Markdown link check, the unit tests of validate-nvx, test_adversarial.py, and test_hosts.py, with Git's bash first on PATH, as on the hosted runners.

Before the rebase, on Linux (WSL), at each commit: rustfmt, both Clippy configurations, and the cli tests; at the head: the full test suite with all features, and the Rust 1.89 check.

Left to CI: Linux after the rebase, the Specula tests, which need fcntl, the macOS check, the guest binaries, and the hypervisor, OpenVMM, and benchmark lanes.

Copilot

🔵 Needs a closer look

Retiring extensive cross-platform regression coverage alongside a persisted-format change warrants human confirmation of the stacked coverage and mutation-test equivalence gates.

No open issues 🎉

What changed in this PR

This PR completes the CLI transcript retirement for #728, building on #825’s structured scenario tests. It replaces recorded-output replay with contract-focused checks and standardizes newly provisioned sandbox configurations on format 7.

Changes:

Replace the golden CLI replay with scenario tests covering errors, launches, lifecycle behavior, and configuration compatibility.
Share a dedicated run-help fixture between Rust and Python tests, and remove obsolete replay coverage tracking.
Write format 7 for all sandboxes while retaining format 1 reads; update documentation accordingly.

File	Description
`scripts/test_nvx_tools.py`	Check Python run help against the shared fixture.
`scripts/test_guest_binaries.py`	Replace the retired fixture reference in source-identity testing.
`scripts/nvx_rs_coverage.py`	Remove the retired replay coverage family.
`nvx-rs/tests/fixtures/state/format-7.json`	Add a format 7 configuration with shares.
`nvx-rs/tests/fixtures/state/format-7-without-shares.json`	Add format 7 with a null export.
`nvx-rs/tests/fixtures/state/format-6.json`	Add an unsupported format 6 fixture.
`nvx-rs/tests/fixtures/state/format-5.json`	Add an unsupported format 5 fixture.
`nvx-rs/tests/fixtures/state/format-4.json`	Add an unsupported format 4 fixture.
`nvx-rs/tests/fixtures/state/format-3.json`	Add an unsupported format 3 fixture.
`nvx-rs/tests/fixtures/state/format-2.json`	Add an unsupported format 2 fixture.
`nvx-rs/tests/fixtures/state/format-1.json`	Add a supported legacy configuration.
`nvx-rs/tests/fixtures/state/format-1-nvx-py.json`	Add a legacy Python configuration for rejection testing.
`nvx-rs/tests/fixtures/state/format-1-network.json`	Add legacy format coverage with networking.
`nvx-rs/tests/fixtures/run-help.txt`	Store shared reference help text.
`nvx-rs/tests/cli/usage.rs`	Check parser errors and help output.
`nvx-rs/tests/cli/openvmm.rs`	Provide command, image, and outcome assertions.
`nvx-rs/tests/cli/one_shot.rs`	Check launch arguments, output, and exit statuses.
`nvx-rs/tests/cli/harness.rs`	Provide isolated roots and structured CLI assertions.
`nvx-rs/tests/cli/formats.rs`	Check supported formats and legacy refusals.
`nvx-rs/tests/cli/egress.rs`	Check policy compilation output and errors.
`nvx-rs/tests/cli/dry_run.rs`	Check dry-run commands and argument round trips.
`nvx-rs/tests/cli/artifacts.rs`	Check artifact listings and guest order.
`nvx-rs/tests/cli.rs`	Replace replay helpers with scenario modules and shared setup.
`nvx-rs/src/host/cli/sandbox/state.rs`	Write format 7 universally and preserve format 1 reads.
`nvx-rs/src/host/cli/sandbox/lifecycle.rs`	Update comments describing persisted configurations.
`nvx-rs/src/host/artifacts.rs`	Update the memory-default test reference in documentation.
`nvx-rs/README.md`	Document scenario tests and configuration compatibility.
`nvx-rs/Cargo.toml`	Enable Windows command-line splitting for tests.
`doc/project-structure.md`	Describe scenario tests and configuration fixtures.
`doc/ci.md`	Document retirement of the replay coverage family.
`.gitattributes`	Remove the retired fixture’s generated-file designation.

_{Review effort: Balanced}

Copilot

🔵 Needs a closer look

Retiring broad cross-platform regression coverage alongside configuration-format changes warrants human confirmation of the Linux and Windows parity evidence.

No open issues 🎉

_{Review effort: Balanced}

Copilot

🔵 Needs a closer look

The cross-platform replay retirement and persisted-format change warrant human confirmation of coverage, mutation-testing, and compatibility evidence.

No open issues 🎉

_{Review effort: Balanced}

nvx.py run parses the options of nvx-rs run before it runs it, so it prints a help of its own, which must be that of nvx-rs run, and which gives the default memory of each guest that guests.py also gives test-microvm. A test of nvx.py compared that help, and the dry runs that nvx.py run recorded for the default of each guest, with the transcripts of tests/fixtures/cli.json, which #728 retires. tests/fixtures/run-help.txt now holds the help of nvx-rs run, as the transcripts recorded it, byte for byte. run_prints_the_help_of_nvx_py_run checks that nvx-rs run prints it, and the test of nvx.py, now test_run_help_is_the_help_of_nvx_rs_run, that nvx.py run prints it too, and that it gives the default memory of each guest of guests.py. The dry runs of the scenario tests check that nvx-rs run boots each guest with that memory. The Testing section of the README and the documentation of Guest::default_memory_mib name the fixture. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

The coverage measurement of #722 compares what each family of golden tests reaches with what the other tests reach, so that the tests that replace a family can show that they reach it too. The family of the CLI's recorded scenarios, nvx_rs_reproduces_the_recorded_scenarios, retires with tests/fixtures/cli.json (#728): the scenario tests of tests/cli/ reach every line and branch of production code that it reached beyond the tests outside the families, on Linux and Windows, and what it alone reached among all the tests was code that only some runs reach. nvx_rs_coverage.py no longer measures the family, whose filter would select no test once the replay is gone, and doc/ci.md notes its retirement. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

tests/fixtures/cli.json held the 315 scenarios that scripts recorded from the Python implementations of nvx.py sandbox and nvx.py run, and nvx_rs_reproduces_the_recorded_scenarios replayed them byte for byte once placeholders replaced the root and volatile values. The scenario tests of tests/cli/ check what callers rely on in their stead (#728): every scenario, and every regression that #722 traces, maps to them; on Linux and Windows, the replay reaches no line or branch of production code beyond the tests outside its family; and cargo-mutants finds no mutant of src/host/cli that the replay catches and the other tests of the CLI miss. The fixture goes, with the replay, its normalization and population helpers, and the mapping of tests/cli/scenarios.rs, whose tables of cases are private again, as is Root::empty(), which only Root::new() calls now. .gitattributes no longer marks the fixture as generated, and the test of the guest binaries' source identity names another file of tests/fixtures, root-tokens.json, which holds the cases of a root's boot tokens rather than behavior recorded from Python, so that no retirement of #722 removes it. The README describes the scenario tests in a section of their own, which replaces Golden scenarios and records how they replaced the transcripts, and its Python parity and Testing sections point to it. #828, which merged after #825 added the scenario tests, changed two messages that they check, and recorded the change in the transcripts alone: an unknown field of an egress policy is quoted as Rust quotes it, "other" where Python wrote 'other', and a policy file that cannot be read is named as it was given, ./ included, where Python normalized its path. The tests of egress compile and of the refusals of run expect these messages, which the transcripts no longer check. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

provision wrote the configuration of a sandbox without shares in format 1, with a null mount, as nvx.py did, and that of one with shares in format 7, with the export plan of its shares. No nvx.py reads either any more: nvx.py runs nvx-rs, and start refuses a configuration without the bindings that only nvx-rs records. #728 asks for the newest format alone. provision now writes format 7 for every sandbox, with export null for one without shares, and start reads format 7, with export a plan or null, and still reads format 1, which earlier builds wrote for sandboxes without shares. A configuration whose format is not the one that its export needs, format 7 with export and format 1 without it, fails as before. The scenario tests expect format 7 of every provision, start a hand-written configuration of format 7 without shares as they start that of format 1, and check that a start refuses a configuration of format 7 without export and one of format 1 with it. The README and the documentation of the state directory describe the formats. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot

🟢 Approval recommended

The focused configuration change preserves legacy reads, explicit scenarios cover the new behavior, and no blocking issues were identified.

No open issues 🎉

_{Review effort: Balanced}

Build on the behavior-preserving specifications of #823 for the retirement half of #730. Remove the six Python-oracle tests of src/host/sandbox/golden.rs and their fixtures, tests/fixtures/sandbox.json and tests/fixtures/sandbox_options.json. Preserve all four accepted launch shapes, eight POSIX and Windows records with identical contracts, as explicit argument, token, and exact-budget regressions in the specification module rather than as fixtures. Remove the launch-profile family from the coverage inventory of scripts/nvx_rs_coverage.py, and update its inventory test to recognize the specification module. #828 and #830 retired the egress-policy and CLI-scenario families on dev, so the inventory and its table in doc/ci.md keep only the Python-semantics family, and doc/ci.md notes this retirement beside that of the CLI family. Update the coverage and mutation documentation, remove the temporary section mapping that #823 kept in the README, and drop the sandbox fixtures from the README's list of Python records. The native input and API changes follow in separate commits. Validation on Windows, rebased on dev: 701 library tests and 51 CLI tests pass with all features, and the coverage-script tests, Ruff, and the Ruff format check pass. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Retire the Python-backed scalar compatibility layer of #730 in favor of native sandbox values while preserving the existing wire contracts. - Remove the public `Int` and `IntError` exports; use `Option<u64>` for the memory limit and `Option<u32>` for the process limit, including the bound that leaves room for the guest's init process. - Share ASCII-digit-guarded `FromStr` parsing for UID and GID, the CLI resource limits, and the limits that sandbox configurations persist, which they keep writing as JSON integers. - Use the canonical native `Uuid` parser while preserving tokens and all valid plan and configuration wire encodings. - Reject malformed limits at the CLI usage stage with clear native errors: `--memory-max` and `--pids-max` fail with `argument-invalid` and exit status 2, naming the accepted ASCII decimal range, instead of Python's `invalid int value` and the later `--pids-max must be positive` refusal with exit status 1. - Quote layer and persisted-plan values with Rust debug formatting. - Remove the now-unused Python integer arithmetic, keep only the test-only historical Python UUID oracle, and narrow the visibility of the remaining private integer helpers. - Update the scenario tests that replaced the CLI transcripts (#825, #830) to match: the usage table expects the double-quoted layer UUID and roles and the native `--memory-max` message, and the `--pids-max 0` case moves from the validation table, whose cases exit with status 1, to the usage table. - Document the input contract and the API change in the README, whose Python parity section no longer lists the sandbox, nor UUIDs among the values that it checks as Python does; #828 already removed the egress policy compiler from that section. Validation on Windows, rebased on dev: 702 library tests and 51 CLI tests pass with all features, and `cargo fmt --all --check` passes. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Retire the Python JSON and surrogate behavior from the sandbox metadata checks for #730. Parse bounded environment files directly as Vec<String> with serde_json and validate ordinary UTF-8 KEY=VALUE strings. Read distro manifests as strict JSON, require the integer format 1, and refuse non-finite numbers and unpaired surrogate escapes, even in unused fields. Preserve the file, entry, and byte limits, the binding of the artifact name and digest, the file seals, and the independent refusal of systemd. Quote sandbox values and paths with Rust debug formatting, and keep the native JSON and I/O causes in diagnostics. Move the now-unused Python JSON lookup and UTF-8 encode-oracle helpers out of production code; the unrelated historical Python family keeps its test-only oracles. Update the focused checks of tests/cli.rs that name these values and paths; the transcripts of tests/fixtures/cli.json, which #830 retired, no longer record them. Document the accepted-input and diagnostic changes. Update the scenario tests that replaced the transcripts (#825, #830) to match. The validation table expects the double-quoted --mount modes and targets, and the serde_json causes of environment files that hold an object or a non-string entry. Its environment-file cases name the file as nvx-rs now quotes it: Root::expand, in the shared harness, gains the {root:?} and {root/RELATIVE:?} placeholders, which quote a path as Rust's Debug quotes it, so that the fragments match the escaped backslashes of Windows paths as well as POSIX paths. Validation on Windows, rebased on dev: 703 library tests and 51 CLI tests pass with all features, `cargo fmt --all --check` passes, and Clippy passes with warnings denied for all targets with all features and for the library without default features. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Complete the kernel-token part of #730 with the existing protocol::cmdline Linux lexer rather than Python's whitespace splitting. Refuse control characters, Unicode whitespace, quotes, and the bytes that Linux treats as whitespace, including the 0xa0 inside a UTF-8 sequence, in entrypoints, workload arguments, and guest targets, including the target of the single directory share of run. Check reserved names after Linux's quote splitting and its aliasing of hyphens and underscores, and also in the view of the guest's shell bootstrap, which splits the expanded command line on IFS whitespace without interpreting quotes, so that a quoted value such as extra="ignored nvx_memory_max=max ignored" cannot override a configured entrypoint, memory limit, process limit, or share; ordinary quoted caller values remain accepted. Trim the caller's command line once, before both reservation checks and rendering, so that the checks see exactly the text that is rendered and leading Unicode whitespace, such as U+2003 EM SPACE, cannot hide a reserved name. Confirm that Linux can still see every generated root parameter, so that -- and unmatched quotes cannot hide the appended boot settings. Keep the guest root parser and the exact 1024-byte budgets for both share-delivery modes and adapter plans. Broaden the properties to quoted and dashed reserved inputs, leading Unicode whitespace, and the known separator bytes, add explicit hidden-root and quoted-override regressions for both share-delivery modes, and make the host-path unit tests native-only. Update the documentation, and the validation table of the scenario tests that replaced the CLI transcripts (#825, #830): a sandbox argument with a space is now refused for its whitespace rather than its spaces. Build the Python compatibility code that only the CLI and the historical tests use for them alone, so that the default-feature library has no dead code, and add `cargo clippy --lib -- -D warnings` to check-nvx-rs so that it stays so. Gate the Python split and strip oracles to tests, and the integer, codec, and Unicode-digit helpers to the cli feature and tests. Since #828 moved the egress policy compiler to serde_json, and the previous commit the sandbox metadata readers, only the CLI's state files use the Python JSON reader and Python strings, so gate the json and text modules whole, as their dumps, shell, and value siblings are. The README's Python parity section now credits the nvx-rs CLI, their only remaining user, with the whitespace, int, and json.loads semantics, and the module documentation of the JSON reader no longer cites the retired egress and distro-manifest readers. Validation on Windows, rebased on dev: 706 library tests and 51 CLI tests pass with all features, and 630 library tests with default features; `cargo fmt --all --check`, Clippy with warnings denied for all targets with all features, for the default-feature library, and for the library without default features, rustdoc with warnings denied, and the Markdown link check pass. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Build on the behavior-preserving specifications of #823 for the retirement half of #730. Remove the six Python-oracle tests of src/host/sandbox/golden.rs and their fixtures, tests/fixtures/sandbox.json and tests/fixtures/sandbox_options.json. Preserve all four accepted launch shapes, eight POSIX and Windows records with identical contracts, as explicit argument, token, and exact-budget regressions in the specification module rather than as fixtures. Remove the launch-profile family from the coverage inventory of scripts/nvx_rs_coverage.py, and update its inventory test to recognize the specification module. #828 and #830 retired the egress-policy and CLI-scenario families on dev, so the inventory and its table in doc/ci.md keep only the Python-semantics family, and doc/ci.md notes this retirement beside that of the CLI family. Update the coverage and mutation documentation, remove the temporary section mapping that #823 kept in the README, and drop the sandbox fixtures from the README's list of Python records. The native input and API changes follow in separate commits. Validation on Windows, rebased on dev: 701 library tests and 51 CLI tests pass with all features, and the coverage-script tests, Ruff, and the Ruff format check pass. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Retire the Python-backed scalar compatibility layer of #730 in favor of native sandbox values while preserving the existing wire contracts. - Remove the public `Int` and `IntError` exports; use `Option<u64>` for the memory limit and `Option<u32>` for the process limit, including the bound that leaves room for the guest's init process. - Share ASCII-digit-guarded `FromStr` parsing for UID and GID, the CLI resource limits, and the limits that sandbox configurations persist, which they keep writing as JSON integers. - Use the canonical native `Uuid` parser while preserving tokens and all valid plan and configuration wire encodings. - Reject malformed limits at the CLI usage stage with clear native errors: `--memory-max` and `--pids-max` fail with `argument-invalid` and exit status 2, naming the accepted ASCII decimal range, instead of Python's `invalid int value` and the later `--pids-max must be positive` refusal with exit status 1. - Quote layer and persisted-plan values with Rust debug formatting. - Remove the now-unused Python integer arithmetic, keep only the test-only historical Python UUID oracle, and narrow the visibility of the remaining private integer helpers. - Update the scenario tests that replaced the CLI transcripts (#825, #830) to match: the usage table expects the double-quoted layer UUID and roles and the native `--memory-max` message, and the `--pids-max 0` case moves from the validation table, whose cases exit with status 1, to the usage table. - Document the input contract and the API change in the README, whose Python parity section no longer lists the sandbox, nor UUIDs among the values that it checks as Python does; #828 already removed the egress policy compiler from that section. Validation on Windows, rebased on dev: 702 library tests and 51 CLI tests pass with all features, and `cargo fmt --all --check` passes. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Retire the Python JSON and surrogate behavior from the sandbox metadata checks for #730. Parse bounded environment files directly as Vec<String> with serde_json and validate ordinary UTF-8 KEY=VALUE strings. Read distro manifests as strict JSON, require the integer format 1, and refuse non-finite numbers and unpaired surrogate escapes, even in unused fields. Preserve the file, entry, and byte limits, the binding of the artifact name and digest, the file seals, and the independent refusal of systemd. Read distro manifests with the strict JSON reader of egress policy files that #828 added, which host::network::policy::json now shares as read_document, with the name that its errors give the root of the document. serde_json's own Value depends on the features of serde_json, which another crate of a build may enable: with arbitrary_precision, it reads an object that holds a number as that number, so that a format of {"$serde_json::private::Number": "1"} would pass the integer check, and it accepts numbers beyond the range of a double in unused fields. The shared reader reads both as it does without the feature. It also refuses an object that repeats a field, as `manifest has duplicate field "format"` or `packages[0] has duplicate field "name"`, and arrays and objects nested more than MAX_JSON_DEPTH, 32, deep, and a manifest that is not UTF-8 now fails as such. Quote sandbox values and paths with Rust debug formatting, and keep the native JSON and I/O causes in diagnostics. Move the now-unused Python JSON lookup and UTF-8 encode-oracle helpers out of production code; the unrelated historical Python family keeps its test-only oracles. Update the focused checks of tests/cli.rs that name these values and paths; the transcripts of tests/fixtures/cli.json, which #830 retired, no longer record them. Document the accepted-input and diagnostic changes. Update the scenario tests that replaced the transcripts (#825, #830) to match. The validation table expects the double-quoted --mount modes and targets, and the serde_json causes of environment files that hold an object or a non-string entry. Its environment-file cases name the file as nvx-rs now quotes it: Root::expand, in the shared harness, gains the {root:?} and {root/RELATIVE:?} placeholders, which quote a path as Rust's Debug quotes it, so that the fragments match the escaped backslashes of Windows paths as well as POSIX paths. The manifest tests refuse that object-valued format as a mismatch, and refuse numbers beyond a double, repeated fields at the root and in a package, nesting beyond the limit, and a manifest that is not UTF-8. A test of the shared reader simulates arbitrary_precision, as its other tests do, and checks that the object stays an object there, and another that read_document names the root of a document in its errors. Validation on Windows, rebased on dev: 705 library tests and 51 CLI tests pass with all features, `cargo fmt --all --check` passes, Clippy passes with warnings denied for all targets with all features, and the Markdown link check passes. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Complete the kernel-token part of #730 with the existing protocol::cmdline Linux lexer rather than Python's whitespace splitting. Refuse control characters, Unicode whitespace, quotes, and the bytes that Linux treats as whitespace, including the 0xa0 inside a UTF-8 sequence, in entrypoints, workload arguments, and guest targets, including the target of the single directory share of run. Check reserved names after Linux's quote splitting and its aliasing of hyphens and underscores, and also in the view of the guest's shell bootstrap, which splits the expanded command line on IFS whitespace without interpreting quotes, so that a quoted value such as extra="ignored nvx_memory_max=max ignored" cannot override a configured entrypoint, memory limit, process limit, or share; ordinary quoted caller values remain accepted. Trim the caller's command line once, before both reservation checks and rendering, so that the checks see exactly the text that is rendered and leading Unicode whitespace, such as U+2003 EM SPACE, cannot hide a reserved name. Confirm that Linux can still see every generated root parameter, so that -- and unmatched quotes cannot hide the appended boot settings. Keep the guest root parser and the exact 1024-byte budgets for both share-delivery modes and adapter plans. Broaden the properties to quoted and dashed reserved inputs, leading Unicode whitespace, and the known separator bytes, add explicit hidden-root and quoted-override regressions for both share-delivery modes, and make the host-path unit tests native-only. Update the documentation, and the validation table of the scenario tests that replaced the CLI transcripts (#825, #830): a sandbox argument with a space is now refused for its whitespace rather than its spaces. Build the Python compatibility code that only the CLI and the historical tests use for them alone, so that the default-feature library has no dead code, and add `cargo clippy --lib -- -D warnings` to check-nvx-rs so that it stays so. Gate the Python split and strip oracles to tests, and the integer, codec, and Unicode-digit helpers to the cli feature and tests. Since #828 moved the egress policy compiler to serde_json, and the previous commit the sandbox metadata readers, only the CLI's state files use the Python JSON reader and Python strings, so gate the json and text modules whole, as their dumps, shell, and value siblings are. The README's Python parity section now credits the nvx-rs CLI, their only remaining user, with the whitespace, int, and json.loads semantics, and the module documentation of the JSON reader no longer cites the retired egress and distro-manifest readers. Validation on Windows, rebased on dev: 708 library tests and 51 CLI tests pass with all features, and 632 library tests with default features; `cargo fmt --all --check`, Clippy with warnings denied for all targets with all features, for the default-feature library, and for the library without default features, rustdoc with warnings denied, and the Markdown link check pass. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Copilot AI balanced review requested due to automatic review settings October 23, 2026 23:42

Copilot started reviewing on behalf of Pedro Henrique Penna (ppenna) October 23, 2026 23:42

View session

Copilot AI reviewed Oct 23, 2026

View reviewed changes

Pedro Henrique Penna (ppenna) force-pushed the agents/issue-728-retire-cli-transcripts branch 2 times, most recently from d334b16 to b77ccbc Compare October 24, 2026 00:23

Copilot AI balanced review requested due to automatic review settings October 24, 2026 00:23

Copilot started reviewing on behalf of Pedro Henrique Penna (ppenna) October 24, 2026 00:23

View session

Copilot AI reviewed Oct 24, 2026

View reviewed changes

Pedro Henrique Penna (ppenna) force-pushed the agents/issue-728-retire-cli-transcripts branch from b77ccbc to e6d11c3 Compare October 24, 2026 00:53

Copilot AI balanced review requested due to automatic review settings October 24, 2026 00:53

Copilot started reviewing on behalf of Pedro Henrique Penna (ppenna) October 24, 2026 00:54

View session

Copilot AI reviewed Oct 24, 2026

View reviewed changes

Pedro Henrique Penna (ppenna) and others added 4 commits October 24, 2026 23:46

Pedro Henrique Penna (ppenna) force-pushed the agents/issue-728-retire-cli-transcripts branch from e6d11c3 to 7b0feef Compare October 25, 2026 03:16

Copilot AI balanced review requested due to automatic review settings October 25, 2026 03:16

Copilot started reviewing on behalf of Pedro Henrique Penna (ppenna) October 25, 2026 03:17

View session

Copilot AI reviewed Oct 25, 2026

View reviewed changes

Pedro Henrique Penna (ppenna) merged commit 50577c9 into dev Oct 25, 2026
45 checks passed

Pedro Henrique Penna (ppenna) deleted the agents/issue-728-retire-cli-transcripts branch October 25, 2026 11:02

Pedro Henrique Penna (ppenna) mentioned this pull request Oct 25, 2026

tests: stage the runtime adapter's images and plan launches for its guest #863

Merged

Pedro Henrique Penna (ppenna) mentioned this pull request Oct 25, 2026

Retire sandbox Python fixtures and enforce native input contracts #826

Merged

This was referenced Oct 25, 2026

ci: the nightly full run is failing #891

Open

ci: test each pull request on top of the latest dev before it merges #895

Open

Navigation Menu

Repository navigation

nvx-rs: retire the golden CLI scenarios - #830

nvx-rs: retire the golden CLI scenarios#830
Pedro Henrique Penna (ppenna) merged 4 commits into
devfrom
agents/issue-728-retire-cli-transcripts

Pedro Henrique Penna (ppenna) commented Oct 23, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Pedro Henrique Penna (ppenna) commented Oct 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Changes

Differences from the issue

Gates

Rebase on dev

Validation

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🔵 Needs a closer look

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🔵 Needs a closer look

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🔵 Needs a closer look

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟢 Approval recommended

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Pedro Henrique Penna (ppenna) commented Oct 23, 2026 •

edited

Loading

Rebase on `dev`