The built-in Python provider scanned PATH for the extension-less token
python3; a Windows PATH contains python3.exe, so a working interpreter was
reported as kernel.unavailable and the semantic golden suite failed on
windows-latest. Expand candidates with PATHEXT (conservative default) and use
the same platform-aware resolver in the conformance test's availability probe.
Also make native failures diagnosable: strip ANSI before matching and emit the
panic/build context as public check annotations, and run Linux tests through
the same annotator. This fixes the Windows native lane and gives the macOS
failure the context needed to diagnose it.
The native matrix only annotated the first matching line, which omitted the
fixture name and expected/actual values. Annotate the panic block and build
error context instead so the failing golden is identifiable from the public
check-run annotations.
The Windows/macOS jobs fail inside cargo test --workspace, but public check
annotations only show the step exit. Tee the output and emit the failing test,
panic or build-error line as a check annotation so the native failure is
diagnosable without repository-admin log access. Behaviour and exit status are
unchanged.
Record the 512b9f5 local disposable-VM run (7/7 plus strict JSON and inert
@knowledge, evidence.json/archive/log digests) and Woodpecker pipeline #16.
The lane now fails if strict-evidence.json cannot be written, so the green run
also proves the evidence path. Fix the cross-platform JSON smoke to tolerate
the expected non-zero strict refusal exit before parsing.
Tested code SHA: 512b9f5. Release decision remains DO_NOT_RELEASE for the
missing remote Windows/macOS native PASS, remote G3 artifact and MEASURED G4.
A green lane without a strict-evidence.json must not count as a gate. The EXIT
trap now turns a generator failure into a lane failure. The self-hosted
workflow exports the archive path in the run step (env written by GITHUB_ENV is
not guaranteed in the env expression context), records the pinned image digest
as an input, and verifies schema/outcome/tested_sha/non-root before uploading.
Also define SOURCE_SHA before the VM overlay commands.
Add tools/ci/release_evidence.py (marqdo.release-gates.v1) plus unit tests:
missing/skipped/FAIL/UNVERIFIED and tested_sha mismatches stay
DO_NOT_RELEASE, performance may be PASS/WARN, unknown gate names are rejected.
The GitHub CI now has a release-evidence job that maps the Linux/schema and
native matrix results onto the manifest and uploads it, and both CI systems
run the evidence-tool tests.
Clear the contradictions the RFC calls out: remove the stale CURRENT.md rows
that claimed DNS/TLS/UDP and fs metadata were absent while the same table said
they were implemented; correct docs/ci.md to state that the GitHub-hosted
Windows/macOS matrix exists (Woodpecker has no such agent); keep the old
convergence plan historical; require the evidence manifest in releasing.md;
and add a docs-consistency guard for these claims.
This is step 6 of the CI/security RFC (B05/B06/B07).
Replace the three-run baseline profile with tools/perf/runner.py: one fixed
release binary, per-child rusage via posix_spawn/wait4, 5 warmups and 20
samples by default, SHA-derived interleaved order, p50/p95/MAD, CPU user/sys
and peak RSS. Every warmup output is normalized and compared with a committed
wire digest before timing; a mismatch is CORRECTNESS_FAIL, never a fast
sample. Short or unapproved-host runs are SMOKE, and baselines from another
platform_id/profile are refused.
Add workloads.json, expected digests for nine existing workloads, a dedicated
fixed-node workflow (manual, self-hosted, requires --cpu-list), and
docs/performance.md. bench.py stays as a compatibility shim. This is step 5 of
the CI/security RFC (B08).
Add tools/ci/strict-evidence.py, which records only measured host/source
facts: tested SHA (git-verified when a checkout exists), source archive
digest, kernel, bwrap/rustc/cargo, uid, cgroup2, runner kind and outcome. The
lane now writes this document on every exit path, refuses a root-only PASS,
and covers the strict @knowledge-inert and forged-marker JSON contracts in
addition to the existing 7/7 adversarial matrix and positive CLI checks.
Rework security-linux.yml to run on trusted master pushes or an explicit
source_sha, bind the checkout to the tested SHA, create a git archive for the
digest, avoid PR execution and upload the evidence artifact. This is step 4 of
the CI/security RFC (B04/B09/B10); the runner must still be an ephemeral VM.
WHY
- The code fixes existed locally without same-SHA release material; the G0..G12
gate matrix, G-C VM evidence, RC artifact and explicit non-claims must be
reviewable together.
- The release-candidate workflow packaged no consumers and did not verify its
own checksum in the clean-install directory.
WHAT
- Record the current Linux/VM evidence at f075c96 and keep remote CI/G-C as
UNVERIFIED; overall decision DO NOT RELEASE.
- Add G-C evidence (7/7 + positive CLI + 47-binary workspace run), RC report
with gate matrix, clean-install and checksum result.
- Update STATUS, COMPATIBILITY, CHANGELOG, README, tests/README, mapping spec
and releasing procedure for the new condition/span/power/Tool behavior.
- Harden release-candidate packaging: include tests/consumers, SOURCE_SHA,
rustc -vV, verify the checksum before extraction, and run independent
Analysis/Trace validation from the extracted archive.
- Record the local convergence note in NEXT_RELEASE_CI_CONVERGENCE.md.
TESTS
- Local fmt/clippy/workspace PASS; G-C VM PASS at f075c96.
- Linux RC: sha256 -c OK and clean-install smoke PASS.
CI_URL
- none: Gitee remote, no GitHub credentials. Windows/macOS and same-SHA G-C CI
remain the blocking gates.
BREAKING_CHANGE
- none in this docs/workflow commit.
SECURITY_SCOPE
- strict remains fail-closed; no new capability or platform claim. G-C is
local-VM evidence, not a CI-signed security URL.
Strict mode now re-executes the CLI inside a Bubblewrap supervisor with
user/mount/PID/net/UTS/IPC namespaces, no-new-privs, all capabilities dropped,
a read-only /project, private /proc and /dev, rlimits, output capture and
wall-clock process-group cleanup. The child verifies PID 1, NoNewPrivs, empty
CapEff, an unshared uid map and loopback-only networking before project code.
- PlatformCapabilities::detect runs a real bwrap probe; no probe, no strict
- strict L1 supports pure Marqdo + Math; Foreign/ABI/worker/kernel/grants/trace
are rejected with exit 2 before sandboxing
- forged MARQDO_STRICT_ACTIVE is rejected; missing bwrap fails closed
- add adversarial helper, 7-test escape/resource matrix, seeded security lane
script and self-hosted security workflow
- executed G-C lane as uid 1000 in a disposable Ubuntu 26.04 VM: 7/7 PASS
plus positive CLI and refusal matrix; evidence in docs/evidence
- ADR-0021, spec/security.md, STATUS/COMPATIBILITY/README/CHANGELOG updated;
ADR-0019 remains for non-Linux/fail-closed rules
- adopt docs/STABLE_RELEASE_IMPLEMENTATION.md as the single next-phase plan
- archive the stale M0.5 status snapshot under docs/history/
- record real local gate results, the red baseline CI, the Windows autocrlf
reproduction and the unresolved macOS/native-CI status
- add COMPATIBILITY.md (protocol versions + honest platform matrix),
CHANGELOG.md, dual license files, docs/releasing.md and docs/security-gate.md
- keep strict fail-closed: document the container EPERM/read-only cgroup probe
and the exact G-C requirements instead of claiming isolation
- document foreign environment binding (ADR-0020) and add a release-candidate
CI job that is gated on all platform jobs
The macOS job failed with no_matching_artifact because the integration test
copied the Linux-only example manifest and built a Linux-named .so. Tests now
generate a manifest for the actual host triple, use the platform dynamic
library name and compiler flags, and never rewrite the tracked example
manifest.
- add a cross-platform no_matching_artifact test that runs without a C compiler
- add real tamper and ABI-major-mismatch rejection tests
- make worker fixtures platform-native; Windows remains explicitly unsupported
(no cc) rather than counted as a pass
- example build scripts emit host-specific build/ fixtures (gitignored)
- CI smoke points at the generated fixtures
- add an independent schema job with a pinned Draft 2020-12 validator that
validates real Analysis v2/Trace v2/Tool artifacts and negative mutations
- tighten the tool schema (SemVer, package-relative entrypoint, non-empty
protocol arrays)
- run cargo test --workspace plus portable CLI/analysis/trace smoke and a
strict-refusal assertion on Windows and macOS
- record honestly that Tool/C-ABI/worker runtime coverage remains Unix-only
- bind traces to the logical project identity and the exact emitted Analysis
digest; every span-bearing event carries module_id
- add host TraceState budgets (100k events / 32 MiB defaults), CLI
--trace-max-events/--trace-max-bytes, terminal trace.truncated events, and a
final serialized byte cap; budgets never change run.status
- disable semantic recording when no trace is requested
- cap worker frames at 8 MiB, align request/result/capability wire caps, make
clean EOF distinct from protocol errors, and mark/kill a worker after a
transport failure instead of reusing it
- add trace-v2 spec/schema, independent trace consumer, CI check and tests
- record the S2/C state in STATUS
- add runnable examples for an unknown-name Tool, multi-export plugin,
external kernel provider and same-name multi-version packages
- add JSON Schemas for Analysis v1, Trace v1 and Tool v1 and validate them in CI
- extend CI with protocol/consumer smoke tests and a Windows/macOS cargo check
- add --kernel provider descriptors and R6 example integration tests
- document the examples and update M0.6 status