Skip to content

Current releases and evidence

Last checked against npm, GitHub, Maven Central, PyPI, crates.io, NuGet and the Go module proxy: 2026-09-30 UTC. TypeScript additionally verified 2026-10-02 UTC at 0.5.0. This is a point-in-time record, not a promise that branch source is published.

This is the canonical current release matrix. Choose a language for installation and a first example. Unreleased branch source must not be represented as part of a published artifact. All versions below remain pre-1.0; the 1.0 milestone is preparation, not publication. Source, package, protocol and corpus versions are distinct; see versioning.

Language Verified version Install Publication evidence
C GitHub c-v0.2.0 Source archive / CMake Record
C# 0.1.0-alpha.1 NuGet Record
Dart 0.1.0 pub.dev Record
Go v0.1.0 Go module Record
Kotlin/JVM 0.1.0 Maven Central Record
Python 0.1.0a2 (alpha) PyPI Record
Rust 0.1.1 crates.io Record
Swift 0.1.0 SwiftPM revision Record
TypeScript npm @deancochran/ftms@0.5.0 npm quickstart Record

The records below retain original hashes, checks and limitations. Existing links and anchors are preserved. They are publication evidence, not instructions for preparing the next release; maintainers should use the release procedure.

npm serves @deancochran/ftms@0.5.0. Tag v0.5.0 identifies reviewed main commit b79bb9e5dfc1af4dbd59ceeddb9e0b26f1d27176, integrated by PR #41.

  • Release run 37030719970 published through npm trusted publishing with provenance. The first public verification attempt timed out while npm processed the accepted upload; the retry verified the existing immutable archive and passed without replacing it.
  • Public archive integrity: sha512-6DhDyn51mj5Ywh7wtv+gcB+GS7O8BwaXVD6oR2poD5ORZ90v2nruDQwJcWg8RQZ4jh3Vrpm7n2VrUginGEGj+g==. The workflow retains the tested archive and installed public-consumer evidence.
  • 594 TypeScript tests pass. CI verifies Node 20, 22 and 24; packed-artifact verification covers ESM/CommonJS loading and declarations, browser resolution, source maps, runtime neutrality and canonical conformance asset identity.
  • The additive decodeIndoorBikeData interface exposes named measurements, raw values and diagnostics. ESM/CommonJS are generated from the same source; existing protocol interfaces and shared corpus versions are unchanged.
  • A public-npm install in Auuki passes its 17 indoor-bike tests and production Parcel build without Parcel/Jest configuration changes. Its full suite retains the same baseline failures (52 tests, 14 suites); 465 tests pass versus 451 at baseline. The proposed integration remains an unmerged upstream PR, not adoption evidence. Minified main bundle increase is 11.44 kB before compression.
  • No live-device, PTS or Bluetooth qualification evidence is implied.

NuGet serves DeanCochran.Ftms 0.1.0-alpha.1, owned by deancochran: https://www.nuget.org/packages/DeanCochran.Ftms/0.1.0-alpha.1.

Terminal window
dotnet add package DeanCochran.Ftms --version 0.1.0-alpha.1
  • Immutable SSH-signed tag csharp-v0.1.0-alpha.1 identifies source commit db1f24a37363c329caf3dce4dcbcf813c6b01a69, integrated by PR #31.
  • Release run 36787466238 passed all gates and published through NuGet Trusted Publishing. Recovery tooling at 99cc6ce5758bc3af6e1863f82769b3627bc524aa (PR #32) corrected a tag-event identity check without changing the tag, source or package version. The earlier attempt stopped before uploading anything.
  • Clean-source Linux/macOS/Windows gates passed; both packed target assemblies (netstandard2.1, net10.0) executed in isolated consumers. Linux x64 NativeAOT passed with the library rooted for complete analysis.
  • Evidence covers 228 codec/raw/inspection cases with 330 comparisons, 63 exact capability cases, and 181,760 measurement layouts in both directions plus 46 sentinel, 47 RFU and 315 prefix cases. No failed/unsupported/skipped cases.
  • Fresh public NuGet consumers compiled and executed both target assemblies; the installed Standard assembly repeated codec, capability and matrix runners.
  • The downloaded package’s NuGet repository signature and deancochran owner were verified using the official client and pinned certificate 1F4B311D9ACC115C8DC8018B5A49E00FCE6DA8E2855F9F014CA6F34570BC482D. Every archive entry other than NuGet’s added .signature.p7s matched the tested upload.
  • GitHub release assets retain the upload, signed public download, symbols package, identity manifest, public-consumer report, corpus evidence and SHA256SUMS.
Artifact SHA-256
Tested unsigned .nupkg upload a22352966722e5176f727313b118aaaa8f429e4d551e6286ca316085e591d54d
NuGet repository-signed .nupkg download 436358d13db7c32557b89f587d619fafd7ef3d79d24edb34aa500732aff1e08d
Submitted .snupkg 31b62075f5763cd842cdea0d65a1c61c1be7e9c623831bc32413ce3c701dd869

The interface remains prerelease. Scope is FullWire, RangeInspection, CapabilityEvidence and NormalizedViews; no fragment planning/assembly, BLE, equipment-control authorization or Bluetooth qualification is claimed. Symbol upload succeeded; asynchronous symbol-server indexing is not separately proven. See the C# guide and release runbook.

The public Go module proxy and checksum database serve github.com/deancochran/ftms/packages/go@v0.1.0. The annotated tag packages/go/v0.1.0 identifies verified clean source bb470e951694bbcaa092fad9a03a3c1c9bb6a757, merged into main by PR #26 at 45e77b89b54d903932ad9b3da3edf929f4ea6324. The tag identifies the reviewed source commit, not the merge commit. No separate repository or root Go manifest is needed.

Terminal window
go get github.com/deancochran/ftms/packages/go@v0.1.0
  • Go CI and project CI passed before merge. Local clean-source gates passed on Go 1.24.0 and 1.27.1.
  • Fresh public consumers on both toolchains downloaded through proxy.golang.org with sum.golang.org enabled, without replace or a workspace. Telemetry, control and capability code ran; downloaded-module tests/build and license verification passed independently of sibling shared fixtures.
  • Public download provenance reports the exact repository, packages/go subdir, tag and source commit above. Module sum: h1:Hmy4+bHNO3r0Tzs0Vpm5E04aWPv9wfSnhywjyDfM7vw=. go.mod sum: h1:ulbbeannQPoraddmBFBr7H6l2cxWlAINU8HEKtTDmoE=. Module zip SHA-256: 907ee6e66d6b2eae75196831917a92735cc958a6eabea84ba20433743503fe9f.
  • All 63 capability cases passed exact whole-report comparison with zero failed/unsupported/skipped cases. Additional gates passed 122 additive raw cases / 216 directional assertions, 9 range-inspection reports, 181,760 structural layouts in both directions, 46 sentinel positions, 47 RFU cases, and 315 incomplete prefixes. Both bounded fuzz targets passed.
  • The GitHub release retains clean-source corpus identities, all four capability-contract hashes, public consumer evidence, matrix/inspection accounting, fuzz results and SHA256SUMS.
  • pkg.go.dev displays the exact v0.1.0 package documentation and executable examples.

Scope is raw bidirectional codecs, range inspection, normalized ranges and static capability interpretation. Normalized measurement views, original normalized codec-v1 comparison, and record planning/assembly remain outside this release. Minimum Go is 1.24; current executed consumer evidence is Linux/amd64. This does not establish real-device interoperability, BLE lifecycle, physical-control permission, or Bluetooth qualification. See the Go guide, coverage and repeatable verification.

  • Registry: https://pub.dev/packages/deancochran_ftms/versions/0.1.0.
  • Source/tag: dart-v0.1.0 identifies 3ca87a685942f942ef1f8c7af5681cf88371bc9b (PR #23).
  • Public archive SHA-256: c3adbd9af2f67a274607b09a6cac997eea4bdac32dbdb7b7cef82e92fc0d50ed.
  • First publication used the authenticated maintainer CLI, as required for pub.dev bootstrap. The downloaded registry archive matched the verified file manifest; a fresh exact-version hosted consumer passed analysis, both public examples, and native compilation/execution.
  • Merged-source CI passed the Linux/macOS/Windows minimum/current SDK matrix, JavaScript/Wasm browser tests and Flutter Android/iOS consumer builds. Local conformance covered all nine corpora, including 181,760 measurement layouts in both directions.
  • GitHub environment pub-dev requires maintainer approval and allows only tags matching dart-v*. The maintainer confirmed pub.dev GitHub automation is set to repository deancochran/ftms, pattern dart-v{{version}}, required environment pub-dev, push events enabled and workflow-dispatch disabled. A future version will provide the first end-to-end OIDC publication evidence.
  • The bootstrap tag’s automatic release run is intentionally cancelled rather than uploading the already-published version again. Do not rerun that upload. Archived 0.1.0 documentation describes its pre-publication candidate state; this release matrix supersedes those historical status statements.
  • Host and consumer checks do not establish real-device BLE interoperability or Bluetooth qualification.

Historical release: npm @deancochran/ftms@0.4.0. The reproducible TypeScript quickstart remains pinned to this version; the current installation table above records the newer release. The status statements below describe verification at the time of this historical release.

PR #5 merged at a32de9c0b108bc55c9dc11752d928e959efd73ac; both version tags identify that commit. Local verification passed 588 TypeScript tests, the full native suite, package consumers and Cortex-M0 compilation. CI, Native C, Publish and Release C all completed successfully for the release commit. The public artifacts were then downloaded and independently verified, not merely inferred from workflow status.

  • npm 0.4.0 is the latest version; registry gitHead matches the release commit. SHA-512 integrity passed, its 42-file layout is preserved, and all seven published source files match. An isolated consumer of the downloaded artifact exercised range inspection and strict normalized numeric bounds successfully.
  • C 0.2.0 archive SHA-256: 3aa60d809f3dcd02634417018d39f325c46a552734f0fe311ba748261914e61f. The downloaded archive passed source-manifest verification and installed, vendored and FetchContent C/C++ consumers plus both repository examples.
  • Remote release consumers passed on Linux, macOS and Windows; Conan/vcpkg recipes passed against the artifact. This is not public registry submission.
  • PR CI initially exposed a hard-coded 0.1 CMake consumer request. The verifier now reads VERSION; the final PR/main/release runs passed. No failed gate was bypassed. GitHub emitted non-fatal Node-action deprecation warnings; migrating the pinned artifact actions is separate maintenance.

Repository examples are not included in the C source archive. Artifact tests copy the repository examples into isolated consumer directories and link them against the extracted/installed archive; they do not claim the archive ships them.

Maven Central now serves io.github.deancochran:ftms:0.1.0: https://repo.maven.apache.org/maven2/io/github/deancochran/ftms/0.1.0/. The signed annotated tag kotlin-v0.1.0 identifies clean source commit 1fdfefb62f5c1b6a6b3e757cf24deb9d5c09560a, not the later PR merge commit. Central deployment 86dfed11-b0e1-4795-a233-60e12e6fc12a reached PUBLISHED. The GitHub release retains the exact Central bundle, signed manifest, public key, publication receipts and signed clean-source conformance reports.

  • Clean-source gates: 87 JUnit tests, all applicable canonical corpora and 181,760 measurement layouts encoded and decoded; all nine PR CI checks passed.
  • Binary, sources, Dokka documentation, POM and Gradle module metadata were signed with OpenPGP fingerprint A355E4B9EBD3FAEC850A84A6B7A6488FE4805547. All ten staged artifacts/signatures matched the prepared manifest before the publish request.
  • All 30 public files (five artifacts, signatures and checksum sidecars) were downloaded and matched the signed manifest. Signatures verified. Independent Kotlin and Java consumers executed and the Android consumer built an APK with FTMS resolution exclusive to the public Maven Central repository.
  • The Portal returned the exact Maven PURL at VALIDATED but empty PURL lists at PUBLISHING/PUBLISHED. A follow-up client fix handles that observed behavior without weakening the pre-publish coordinate/staged-byte checks. It has 18 passing offline publishing-safety tests. No artifact or tag was replaced.
Artifact Verified SHA-256
Binary JAR c32c3d4bf2533c0d5ff20cc30778973e0c1fd9568a7e75c505a6db0cbaf6a650
Sources JAR cbbc3b3363125c7501ff45eb0780f6bfe21747988ccd015452c7bf126afaa8b7
Documentation JAR b940124bbf6684b37b2333a8178ea4250a8c0243b1af688f8bae99a23ce4771a
Central upload bundle f56b8db378ab1efb7242126fbbed42fef7d1761c9a2316e735d8c2c3a126eb25

The repeatable, explicit prepare → upload → publish → verify process and credential/key maintenance are documented in the Kotlin release runbook. Kotlin package version, FTMS protocol revision and corpus identity remain independent. This is host/artifact and Android-build evidence, not a live Kotlin BLE or Android-runtime test, physical accuracy result, or Bluetooth qualification.

Future Kotlin releases are automated by Release Kotlin/JVM after a maintainer pushes the matching signed annotated kotlin-vVERSION tag on a reviewed main commit. Merging a version/changelog update does not publish. The tag workflow runs the full gate, publishes to Central and verifies public consumers before completing its GitHub release. The persistent maven-central environment holds the existing credentials/signing material; it does not require repeating local integration setup for each version. Existing published versions are verified without another upload. See the runbook for recovery and rotation.

Verified against public GitHub release assets and a fresh public-tag SwiftPM consumer on 2026-09-30. Implementation PR #7 and release-hardening PR #8 are merged. The annotated tag swift-v0.1.0 identifies a18009d6e9892d92bba00e3c6c6388a9fbc0f5c8. No existing npm/C tag or artifact was changed and no separate repository was created.

  • Release Swift run 36662538259 succeeded, including tag identity, clean-source/main ancestry, complete corpus identity checks, Linux/macOS native verification and public tag-pinned consumers.
  • Both hosts passed 16 native tests and all 282 canonical fixture IDs with zero failures, skips, unsupported or unresolved cases. The shared measurement matrix passed 181,760 layouts, 46 sentinels, 47 RFU cases and 315 incomplete prefixes. A deterministic 2,080-payload malformed-input exercise did not trap.
  • Linux used Swift 6.0.3; macOS used Apple Swift 6.1.2, Xcode 16.4 (16F6). Installed-consumer release builds passed for macOS 13, iOS 16, tvOS 16, watchOS 9 and visionOS 1 deployment targets. The reports retain resolved Git identities, SDK inventory, toolchains and canonical schema/vector/contract hashes.
  • Downloaded all four public JSON reports and verified SHA256SUMS and every reported canonical-file hash against the tagged source. A second fresh public tag consumer on Linux compiled and ran after publication, resolving the exact commit above.
  • Early gates exposed a local consumer’s missing macOS minimum, hidden artifact upload filtering, incomplete older Xcode platform components and Git line-ending attributes that made clean checkouts appear dirty. All were fixed before the tag. The generated Kotlin launcher bytes were preserved unchanged. CI now requires clean evidence; no failed gate was bypassed. Pinned artifact actions still emit non-fatal Node-runtime deprecation notices.
Public evidence asset Verified SHA-256
swift-linux-verification.json a96c426d6e98df20263bdcbdcd3ee9f8ff70f61b6a48f9b34f63c772d4c0b9f7
swift-linux-consumer.json 0fb205e5ad9854e0849c95d6e7a126f2b8a6298e920253fa5123c86473d3f59c
swift-apple-verification.json 0587679c91659c967378ecd4c341f6b5d4ec479904e4b7b6edb10c48c3a1640c
swift-apple-consumer.json f5b0705208f62c49d1e31399e9f4a1370b93b59de90adfba9be47931aeeecf0d

Use .package(url: "https://github.com/deancochran/ftms.git", revision: "swift-v0.1.0") and product FTMS; use the full commit above for an immutable pin. Normal SwiftPM version requirements are intentionally unsupported because ordinary v* tags in this repository identify npm releases. This is native compiler, host-runtime and SDK-build evidence, not Swift BLE device interoperability, runtime coverage of every Apple OS version, physical accuracy or Bluetooth qualification. See the Swift release runbook.

PyPI serves deancochran-ftms==0.1.0a2 at https://pypi.org/project/deancochran-ftms/0.1.0a2/. Published and verified on 2026-09-30 from annotated tag python-v0.1.0a2, identifying clean source commit 82d978fa980fc4c5629fbdd855af42126f49bcef. Implementation PR #25 is merged. The release workflow passed verification, published through the approved pypi environment using Trusted Publishing and attestations, and completed public-artifact verification.

Public artifact Verified PyPI SHA-256
deancochran_ftms-0.1.0a2-py3-none-any.whl b9a1feb44140f52c32ecfcb1841d41cfa05ea0627b4361706f6934179cdc60a8
deancochran_ftms-0.1.0a2.tar.gz 5261fa144e0f58ebe95d819ad276a269492abb22cb12c9248e42a774e04972bd

Both public files were downloaded and matched the exact verified build hashes. Independent non-editable wheel and sdist consumers passed on Python 3.11.16, including feature/control/measurement codecs, capability enum exports and explicit signed-tenths resistance evaluation. The workflow retains python-public-evidence-82d978fa980fc4c5629fbdd855af42126f49bcef, whose public-package-verification-report.json reports complete: true and no errors.

The release passed 169 tests on each of Python 3.11 and 3.14, all scoped codec corpora, the 181,760-layout measurement matrix, and 63/63 exact capability cases with zero failed, unsupported or skipped cases. Capability schema version 1 and comparison inputs consumed by this release have these SHA-256 identities:

Capability corpus / comparison input SHA-256
shared/conformance/capabilities/v1/schema.json 1a23dd523896d41b6aa115eea906e6f899a9cfcc8008a87d133ba8c51409ef26
shared/conformance/capabilities/v1/vectors.json 90a9b85e735455515c36fc089fa786bd928e217e81cf95f5ccef67c0d479d3dd
shared/conformance/capabilities/README.md e844292d9a916aa63db9d1f6d22de5525c1923e3584afbcc5013d93d374e6a6a
shared/protocol/capability-discovery.md 541ddd5950999bcd048808fd6eff578dc0a7031eaaa0dc988f8711003ecd488f

The alpha adds CapabilityEvidence to existing FullWire raw codecs, RangeInspection and selected NormalizedViews. It does not add a normalized control encoder, record planning/assembly, BLE transport or execution authority. These are host/artifact results, not device interoperability or qualification.

Historical release evidence; superseded by 0.1.0a2 above.

PyPI serves deancochran-ftms==0.1.0a1 at https://pypi.org/project/deancochran-ftms/0.1.0a1/. The annotated tag python-v0.1.0a1 identifies clean source commit f556d9f9d6e5fc253c1ac9bd4564ea30ad97c26d. The tag’s Python verification and release run completed successfully and published the verified wheel and source distribution through PyPI Trusted Publishing.

Public artifact Verified PyPI SHA-256
deancochran_ftms-0.1.0a1-py3-none-any.whl 3cd84fee28c3e8b2968d4fcbf9451c343b1b72e949e22797fc52d2deae8bf4b0
deancochran_ftms-0.1.0a1.tar.gz 0f434f7e3b14f499815367f028cb7af1744d9bc5156c80f364cc21b7a1190de4

The release is an explicitly selected pre-release with an evolving interface. It provides bidirectional raw Feature, range, control, measurement and status codecs plus selected normalized views and range inspection. It does not provide static capability evaluation, BLE integration, lifecycle policy or execution permission. Host and package verification are not real-device evidence. The release workflow now includes a post-publication public-artifact hash and isolated consumer job for future Python tags. A read-only rerun downloaded both artifacts, matched the hashes above and executed independent isolated wheel and sdist consumers on Python 3.11.15; both completed successfully.

crates.io serves ftms 0.1.1, with docs.rs documentation. Annotated tag rust-v0.1.1 identifies 97885265d28b22807499ada88882cda05f51ff9a, the reviewed merge of PR #34. Release run 36790652901 passed identity, Linux MSRV/stable, macOS, Windows and publication gates.

  • The release adds allocation-free normalized range, control-request/response and Machine Status projections. Normalized codec-v1 accounts for 97/97 cases; capability-v1 accounts for 63/63 reports, with no unsupported/skipped cases.
  • The tested .crate, freshly downloaded crates.io archive and GitHub release asset are byte-identical, SHA-256: 11186b0356d9cc0396d5806f7ec08191ecaeb0a9c80fa2350b580b4f3f3313fe.
  • The clean-source release record identifies Rust/Cargo 1.85.1 and an executed isolated archive consumer. registry-evidence.json records the fresh public registry consumer as passed for the same version, commit and digest.
  • The GitHub release retains the crate, release.json, registry-evidence.json and SHA256SUMS. All checksum entries were independently verified after publication. Registry metadata confirms the same digest and that the version is not yanked.
  • release.json records every corpus/schema/comparison-contract hash. Codec-v1 schema/vector hashes remain e6d976172a32e3124602d17e46ed5fbab337f4f1d6535aec85a83069637268e0 and 9b5b61353b191179cc629d8c459b2a185b9e03b7ce5f3f51b715520162e8a5c7; comparison contract shared/conformance/README.md is 4ee407ecca3cdce77289ff15920bd831dc6687e5dff0ac9c0a8edf39d6034cc9. The capability contract’s documentation-only status update is pinned separately as 0eb4267e0a1c52c99738d6ccc22c9d62cb980d4d66b730ffc8ef2fd648e7c183.

The release notes quote the tagged changelog’s pre-publication wording; the successful public verification above supersedes that preparation status. No tag or immutable package was changed to update this documentation. Cortex-M0 evidence is compilation only; real-device operation and Bluetooth qualification remain unverified.

Historical release evidence; superseded by 0.1.1 above.

crates.io serves ftms 0.1.0. The annotated tag rust-v0.1.0 identifies 58b9add0addea516769c6f01d4b117a1b02110e1; its release archive SHA-256 is fbb9cf806ba89441ef44bab25b65c7caab405c83efe8ededcfd42cc04011f86f. Release Rust run 36763795982 succeeded, and docs.rs is live. This verifies the 0.1.0 artifact, not the later range/control/status normalized projections and expanded coverage in 0.1.1.

The released artifact provides FullWire raw codecs, RangeInspection, capability interpretation, normalized Feature/measurement views and record planning/assembly. The newer range/control/status normalized projections and 97-case runner evidence belong to 0.1.1. BLE, device runtime and control authorization are not provided. See the package’s verification record for exact host, corpus, Cortex-M0 compile-only and packed-consumer evidence.

Exact versions and installation links are in Current verified releases. The following summarizes the boundaries of the records above, not additional releases.

Language Verified consumer evidence Not established by that evidence
C Installed C client and passive replay Public vcpkg/Conan registry availability
C# Both public target assemblies executed in isolated consumers Stable API, Unity/MAUI/device runtime or BLE qualification
Dart Fresh hosted consumer Physical-device or Bluetooth qualification evidence
Go Public proxy/checksum database and isolated consumers Measurement normalization, record planning/assembly or device evidence
Kotlin/JVM Public-artifact Kotlin/Java execution and Android APK build Kotlin Multiplatform, BLE transport or Android runtime/device evidence
Python Public wheel/sdist consumers including capability evaluation Stable interface or live-device evidence
Rust Public-registry installed consumer and docs.rs BLE/device runtime or qualification evidence
Swift Public revision-pinned Linux/macOS consumers and Apple SDK builds BLE integration or Apple-device runtime evidence
TypeScript Current quickstart; separate client preserves a historical compatibility baseline Native implementations as npm exports

Start with the released TypeScript package

Section titled “Start with the released TypeScript package”

Follow the installed-package quickstart for installation and execution commands.

The current-release quickstart asserts normalized metrics, truncation behavior, capabilities, ranges and byte conversions without communicating with equipment. The separate typescript-client example preserves the historical npm 0.2.0 baseline.

The C example uses find_package(ftms CONFIG REQUIRED) against an installed artifact, not private source paths. The source-package verifier builds/installs the archive, moves its prefix and runs the example from an isolated copy. The public archive is at https://github.com/deancochran/ftms/releases/tag/c-v0.2.0. Source availability does not establish package-manager registry publication.

Historical baseline identity and next actions

Section titled “Historical baseline identity and next actions”
  • Previous npm 0.3.0 and C 0.1.0 releases identify 74f1552959d96755f38eac42f6999a5b04088b2f.
  • npm 0.3.0 integrity and all seven included source files matched that checkout.
  • Previous C 0.1.0 archive SHA-256 is 3dc61329a2883f88a7828cff77b282961c9e91680ecae0ca5100622e9d8e6924; its checksum asset and ten source/header files matched the reviewed checkout.
  • CI, Native C, Publish and Release C runs for that commit succeeded. This does not establish environment protection settings or device compatibility.
  • Further changes require separate review, verification and explicit release approval. Updating this record does not publish a new version.
  • vcpkg/Conan registry submissions are separate external actions after a real immutable artifact exists. Local recipe tests are not registry publication.

See C release runbook and real-equipment test procedure. The repository includes a limited passive KICKR CORE pilot. It used the earlier local 0.1.0-based C installation, not the newly published 0.2.0 artifact. Publication and host replay do not expand that device evidence. Do not advertise compatibility with specific equipment until reviewed evidence names the actual model, firmware, platform and installed package version.