Repository navigation
Releases
v0.1.41
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
Full Changelog: v0.1.40...v0.1.41
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.41ghcr.io/ejackson-vp/sandbox-guest:v0.1.41
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.41Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.41 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.40
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
Full Changelog: v0.1.39...v0.1.40
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.40ghcr.io/ejackson-vp/sandbox-guest:v0.1.40
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.40Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.40 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.39
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- ci: bump the actions group with 3 updates by @dependabot[bot] in #265
- deps: bump node from
8996bb5to6bd9ab4by @dependabot[bot] in #262 - deps: bump alpine from 3.21 to 3.24 in /guest by @dependabot[bot] in #263
- deps: bump golang.org/x/crypto from 0.58.1-0.20261009181718-2f1f834330b2 to 0.59.0 in /guest by @dependabot[bot] in #261
- deps: bump the dev-dependencies group with 4 updates by @dependabot[bot] in #264
- feat: list the resource in the Bazaar, and ship an MCP server by @ejackson-vp in #278
Full Changelog: v0.1.37...v0.1.39
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.39ghcr.io/ejackson-vp/sandbox-guest:v0.1.39
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.39Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.39 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.37
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- fix(api): an allowlisted create no longer sends a null settlement nonce by @ejackson-vp in #276
- release: pin 0.1.37 by @ejackson-vp in #277
Full Changelog: v0.1.36...v0.1.37
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.37ghcr.io/ejackson-vp/sandbox-guest:v0.1.37
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.37Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.37 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.36
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- deploy: the settlement RPC reaches every workload that validates payments by @ejackson-vp in #274
- release: pin 0.1.36 by @ejackson-vp in #275
Full Changelog: v0.1.35...v0.1.36
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.36ghcr.io/ejackson-vp/sandbox-guest:v0.1.36
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.36Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.36 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.35
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- release: pin 0.1.34 by @ejackson-vp in #267
- site: the product is throwbox by @ejackson-vp in #268
- site: drop the small print, and stop pointing at evidence by @ejackson-vp in #269
- release: pin the promoted 0.1.34 pair by @ejackson-vp in #270
- fix(api): sign CDP tokens with the key's algorithm, not a constant by @ejackson-vp in #272
- sdk: a mainnet payer, behind an acknowledgement by @ejackson-vp in #271
- release: pin 0.1.35 by @ejackson-vp in #273
Full Changelog: v0.1.34...v0.1.35
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.35ghcr.io/ejackson-vp/sandbox-guest:v0.1.35
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.35Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.35 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.34
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- release: pin the promoted 0.1.33 guest by @ejackson-vp in #259
- docs: re-score the gate, and stop the ORR recommending the wrong stop by @ejackson-vp in #260
- deploy: keep the tunnel's ingress rules in Git by @ejackson-vp in #266
Full Changelog: v0.1.33...v0.1.34
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.34ghcr.io/ejackson-vp/sandbox-guest:v0.1.34
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.34Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.34 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.33
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- docs: record the 0.1.32 isolation run and point the site at it by @ejackson-vp in #242
- config: fail closed on the insecure-dev flag, and stop retyping bounds by @ejackson-vp in #244
- refactor(protocol): declare every /internal answer once, in shared by @ejackson-vp in #246
- deploy: stop shipping a config the controller refuses to start by @ejackson-vp in #245
- test(contract): the contract suite needs no guest image by @ejackson-vp in #250
- fix(store): fence only the operation a session's gate owns by @ejackson-vp in #247
- docs(adr): record the capability key ring and supersede 0001 by @ejackson-vp in #248
- docs(runbook): what to do when a refund is stuck by @ejackson-vp in #249
- fix: jitter every retry loop, and read the Retry-After we already send by @ejackson-vp in #251
- refactor: make a swapped relay/proxy endpoint fail to compile by @ejackson-vp in #253
- lint: ratchet function length, not just file length by @ejackson-vp in #252
- refactor(controller): one module per payment route, and the tests that found three leaks by @ejackson-vp in #255
- fix(capacity): an allocated reservation names a session, or it has a deadline by @ejackson-vp in #256
- release: name the guest version, because it can lag by @ejackson-vp in #257
- release: pin 0.1.33 by @ejackson-vp in #258
Full Changelog: v0.1.32...v0.1.33
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.33ghcr.io/ejackson-vp/sandbox-guest:v0.1.33
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.33Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.33 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
Isolation evidence for the promoted pair
The skip above is about CI: the preflight needs a KVM runner this cluster does not have. The suites were run afterwards against the digests this release promoted, on the real cluster:
| Control | 0.1.33 sha256:4804da2e… |
| Guest | 0.1.33 sha256:933a53e0… |
| Result | 23 passed, 1 skipped (the X-Forwarded-For case; this deployment sets TRUST_PROXY_HOPS=1) |
Recorded in docs/isolation-evidence.md, and the gate row in security-review/RELEASE_GATE.md names these digests.
What the tag alone deploys
The guest image stamps BUILD_COMMIT into a label, so the digest of a release cannot exist until after the commit that pins it. This tag therefore pins guest 0.1.32, and the commit that follows promotion moves the overlay to guest 0.1.33.
Deploying by tag alone gives you control 0.1.33 with the previous guest. Deploy the commit after it, or pin the guest digest explicitly, if you want the reconnect jitter this release added to the guest.
v0.1.32
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- feat: a create kill switch that does not stop live sessions by @ejackson-vp in #224
- feat: Postgres PITR to an object store, with a drill that was actually run by @ejackson-vp in #228
- docs(security): close F-19 and the Kata row in the gate by @ejackson-vp in #229
- docs(site): beta copy that states what is and is not proven by @ejackson-vp in #232
- test(security): let the probe pod pull the image the cluster runs by @ejackson-vp in #230
- cleanup(shared,sdk,cli): one implementation each, and a smaller SDK by @ejackson-vp in #235
- cleanup(api): one definition per repeated answer by @ejackson-vp in #237
- test(security): assert a usable Retry-After, not a boolean by @ejackson-vp in #238
- cleanup(controller): one reader per shape, one decision per question by @ejackson-vp in #236
- cleanup: delete nothing that is used, and correct the docs that drifted by @ejackson-vp in #234
- fix: a tag cannot pin a different release, and bumping is one command by @ejackson-vp in #231
- fix(test): make SIGSTOP observable before a frame is forwarded by @ejackson-vp in #240
- test(property): generate unique settlement nonces by @ejackson-vp in #239
- docs(orr): record the one alert delivery that happened by @ejackson-vp in #233
- release: allow a lagging guest, and pin 0.1.32 by @ejackson-vp in #241
Full Changelog: v0.1.31...v0.1.32
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.32ghcr.io/ejackson-vp/sandbox-guest:v0.1.32
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.32Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.32 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.
v0.1.31
⚠️ Isolation suites were not run for this tag
The Cilium packet-level suite and the Kata boot probe did not execute, so
tenant isolation is unverified and the control and guest image digests
were not verified by them. This release continues only because a
workflow_dispatch run set isolation_skip=true (acknowledgement none):
a tag push cannot take this path.
Set SANDBOX_ISOLATION_RUNNER and re-run the release without the skip
input. See docs/release-process.md#isolation-gate.
What's Changed
- test: pin the reconciler's decision rules with property tests by @ejackson-vp in #198
- test: check the PodSpec over the wire, not the returned name by @ejackson-vp in #199
- test: fuzz the create, internal payment and controller-client surfaces by @ejackson-vp in #200
- test: pin capacity, lifecycle and warm-pool invariants by @ejackson-vp in #201
- test: boot the real API and controller in process by @ejackson-vp in #202
- test: count the fuzz targets in the badge by @ejackson-vp in #203
- fix: answer a rejected body the same way on both API surfaces by @ejackson-vp in #204
- test: fuzz the internal session routes and the refund worker by @ejackson-vp in #206
- test: pin the scheduler tick and the pool allocation policy by @ejackson-vp in #207
- test: give capacity fixtures distinct node ids by @ejackson-vp in #208
- test: add mutation testing for the three money modules by @ejackson-vp in #209
- ci: run the packet-level isolation suite per relevant pull request by @ejackson-vp in #205
- test: measure the Go guest and the fuzz corpora by @ejackson-vp in #210
- test: kill the security-relevant payment mutants by @ejackson-vp in #211
- test: drive the SDK transfer loops through a fake guest by @ejackson-vp in #212
- test: point the blind fuzz targets at guest code, and prune leaked databases by @ejackson-vp in #213
- fix: refuse an incomplete accepted block instead of throwing, and turn on two gates by @ejackson-vp in #214
- security: close the launch blockers for a paid public beta by @ejackson-vp in #215
- docs: add the ordered launch runbook by @ejackson-vp in #216
- docs(security): rebase the review on the current revision and re-gate by @ejackson-vp in #217
- feat(deploy): order tenants below the control plane, and remove private deployment references by @ejackson-vp in #218
- fix(payments): prove the settlement transfer, not just the nonce by @ejackson-vp in #219
- fix: make the two easy-to-leave-off states visible by @ejackson-vp in #220
- fix(controller): record settled-but-unentitled debt from the chain by @ejackson-vp in #222
- docs: operational readiness review, and two alerts that could never fire by @ejackson-vp in #223
- test: attack the payment path, and fix the settlement record it broke by @ejackson-vp in #225
- test: produce the Kata isolation evidence the gate was missing by @ejackson-vp in #227
- test: raise api and sdk coverage, and ratchet the floors by @ejackson-vp in #226
Full Changelog: v0.1.30...v0.1.31
Images
Multi-arch (linux/amd64, linux/arm64), published to GHCR:
ghcr.io/ejackson-vp/sandbox-control:v0.1.31ghcr.io/ejackson-vp/sandbox-guest:v0.1.31
These tags were created only after both images passed the Trivy
gate. Each image carries an SBOM and SLSA build provenance as OCI
attestations:
docker buildx imagetools inspect ghcr.io/ejackson-vp/sandbox-control:v0.1.31Verifying provenance
The *.sigstore.json assets are keyless Sigstore bundles produced by
this workflow, covering exactly the digests above. Verify the image
itself:
gh attestation verify \
oci://ghcr.io/ejackson-vp/sandbox-control:v0.1.31 \
--repo ejackson-vp/sandboxAdd --bundle-from-oci to check the attestation stored in the
registry instead of the one on GitHub.
*.intoto.jsonl is the in-toto statement carried inside the bundle,
for SLSA tooling that consumes statements rather than bundles.
The guest image is what runs inside the microVM and must match the
cluster architecture. CVE results for this tag are in the Security
tab (trivy-control, trivy-guest).
Staging images from the build (staging-<run-id>-<sha>) are not
release artifacts; see docs/release-process.md for cleanup.