Public mode

No access to your tenant or devices. Every item links to its source. What this means

Sources need attention Customize Public modeDiscuss ChangeIntel on Discord

sealed-secrets /v1/rotate can widen sealing scope to cluster-wide via attacker-controlled template annotations

CVSS 3.1 4.9 MEDIUMEUVD CVSS 3.1 4.9 CVSS 3.1 · [email protected] · Secondary 4.9 MEDIUMCVSS_V3 Publisher rating MODERATEFixed by github.com/bitnami-labs/sealed-secrets 0.36.0osv · severity unknownFixed by github.com/bitnami-labs/sealed-secrets 0.36.0KEV status unknown

cve-program · CVE-2026-22728

Published

Bitnami Sealed Secrets is vulnerable to a scope-widening attack during the secret rotation (/v1/rotate) flow. The rotation handler derives the sealing scope for the newly encrypted output from untrusted spec.template.metadata.annotations present in the input SealedSecret. By submitting a victim SealedSecret to the rotate endpoint with the annotation sealedsecrets.bitnami.com/cluster-wide=true injected into the template metadata, a remote attacker can obtain a rotated version of the secret that is cluster-wide. This bypasses original "strict" or "namespace-wide" constraints, allowing the attacker to retarget and unseal the secret in any namespace or under any name to recover the plaintext credentials.

Read the publisher's advisory

euvd · EUVD-2026-8795

Published

Bitnami Sealed Secrets is vulnerable to a scope-widening attack during the secret rotation (/v1/rotate) flow. The rotation handler derives the sealing scope for the newly encrypted output from untrusted spec.template.metadata.annotations present in the input SealedSecret. By submitting a victim SealedSecret to the rotate endpoint with the annotation sealedsecrets.bitnami.com/cluster-wide=true injected into the template metadata, a remote attacker can obtain a rotated version of the secret that is cluster-wide. This bypasses original "strict" or "namespace-wide" constraints, allowing the attacker to retarget and unseal the secret in any namespace or under any name to recover the plaintext credentials.

Read the publisher's advisory

nvd · CVE-2026-22728

Published

Bitnami Sealed Secrets is vulnerable to a scope-widening attack during the secret rotation (/v1/rotate) flow. The rotation handler derives the sealing scope for the newly encrypted output from untrusted spec.template.metadata.annotations present in the input SealedSecret. By submitting a victim SealedSecret to the rotate endpoint with the annotation sealedsecrets.bitnami.com/cluster-wide=true injected into the template metadata, a remote attacker can obtain a rotated version of the secret that is cluster-wide. This bypasses original "strict" or "namespace-wide" constraints, allowing the attacker to retarget and unseal the secret in any namespace or under any name to recover the plaintext credentials.

Read the publisher's advisory

osv · GHSA-465p-v42x-3fmj

Published

This report shows a scope-widening issue in the rotate (re-encrypt) flow: the output scope can be derived from untrusted `spec.template.metadata.annotations` on the input sealed secret. If a victim sealed secret is strict- or namespace-scoped, an attacker who can submit it to the rotate endpoint can set `sealedsecrets.bitnami.com/cluster-wide=true` in the template metadata and receive a rotated sealed secret that is cluster-wide, enabling retargeting (`metadata.name`/`metadata.namespace`) and unsealing to recover the victim plaintext. ## Relevant Links (Pinned) - Rotate handler uses `NewSealedSecret(..., secret)` after unsealing: https://github.com/bitnami-labs/sealed-secrets/blob/946bc048f3407117c837da6e4300686522d4c4eb/pkg/controller/controller.go#L560-L606 - Scope derivation reads secret annotations (`SecretScope`): https://github.com/bitnami-labs/sealed-secrets/blob/946bc048f3407117c837da6e4300686522d4c4eb/pkg/apis/sealedsecrets/v1alpha1/sealedsecret_expansion.go#L112-L122 ## Root Cause The rotate flow unseals the input sealed secret to a `Secret`, then reseals using `NewSealedSecret(..., secret)`. Because `SecretScope(secret)` is computed from secret annotations, and unsealing applies `spec.template` metadata onto the unsealed secret, an attacker can influence the scope of the rotated output by mutating template annotations on the rotate input. ## Attack Path 1. Attacker obtains a victim `SealedSecret` object (for example via read access to resources or logs) and can submit it to the controller rotate endpoint. 2. Attacker sets `spec.template.metadata.annotations.sealedsecrets.bitnami.com/cluster-wide=true` (and optionally retargets name/namespace fields). 3. Rotate returns a resealed, cluster-wide sealed secret that is no longer bound to the victim name/namespace. 4. Attacker unseals the rotated output in their chosen namespace/name to recover the victim plaintext. ## Proof of Concept Setup + run: ```bash unzip poc.zip -d poc cd poc make test ``` Canonical output (excerpt): ``` [CALLSITE_HIT]: pkg/apis/sealedsecrets/v1alpha1/sealedsecret_expansion.go:112 SecretScope [PROOF_MARKER]: scope_widened=true rotated_scope=cluster-wide ``` Control output (excerpt): ``` [NC_MARKER]: scope_widened=false strict_scope_preserved=true ``` ## Fix Accepted When Rotate preserves the original sealing scope and does not allow scope widening based on untrusted template metadata; strict or namespace-wide inputs cannot produce cluster-wide outputs. [poc.zip](https://github.com/user-attachments/files/25080027/poc.zip) [PR_DESCRIPTION.md](https://github.com/user-attachments/files/25080028/PR_DESCRIPTION.md) [attack_scenario.md](https://github.com/user-attachments/files/25080029/attack_scenario.md)

Read the publisher's advisory

osv · GO-2026-4565

Published

Sealed Secrets for Kubernetes: Rotate API Allows Scope Widening from Strict/Namespace-Wide to Cluster-Wide via Untrusted Template Annotations in github.com/bitnami-labs/sealed-secrets

Read the publisher's advisory

Affected versions, risk and source history

Related advisories and sources

Related changes

Linked by CVE identifier

No related changes are available here.

ChangeIntel

An IT change radar: releases, security, known issues, retirements, documentation changes, and service status from public sources. Every item links to supporting evidence; dates and statuses can change after they are read.

Sources read 16 Nov 11:47 UTC · 274 of 275 readable · 250 fully read · 20 waiting for GitHub's hourly limit · documentation 469/469 current