GHSA-7M8X-QG2J-4M3V
Vulnerability from github – Published: 2026-06-30 18:12 – Updated: 2026-06-30 18:12Summary
The Fission MessageQueueTrigger (MQT) scaler controller exposed two privilege-escalation primitives to any subject able to create MQTs in a namespace.
Details
1. Secret materialization. getEnvVarlist in pkg/mqtrigger/scalermanager.go read the Secret named in Spec.Secret using the controller's cluster-wide secrets/get RBAC and emitted each key as a literal EnvVar.Value, copying
the plaintext secret content into the connector Deployment's pod template. A subject holding messagequeuetriggers/create but not secrets/get could exfiltrate any Secret in the namespace by pointing an MQT at it.
2. PodSpec injection. Spec.PodSpec was merged into the controller-built connector PodSpec via util.MergePodSpec with no allowlist on which fields could come from the user. An MQT could substitute Containers[].Image (run any
image), override Command/Args, inject Env, add VolumeMounts + Volumes, override ServiceAccountName, and set HostNetwork/HostPID/HostIPC — turning messagequeuetriggers/create into effective deployments/create with an
arbitrary image and service account.
Impact
A tenant with only messagequeuetriggers.fission.io/create in a namespace could read any Secret in that namespace and run an arbitrary container image under an arbitrary service account, escalating well beyond their intended RBAC.
Fix
Fixed in #3367 and released in v1.24.0.
getEnvVarlistnow emitsEnvVar.ValueFrom.SecretKeyRefso the connector pod resolves values at start time under its own service account. The secret values are never written into the Deployment object and never logged.- A new allowlist,
MergeAllowedPodSpecFields(pkg/executor/util/merge_allowlist.go), accepts onlyNodeSelector,Tolerations,Affinity,RuntimeClassName, and per-containerResources. All other user-supplied fields are dropped at the controller layer, and the validating webhook rejects every populated non-allowlisted field with a clear error. The webhook and the merge helper share a single canonicalDisallowedPodSpecFieldsenumeration so they cannot drift.
Behavioural change
MQT authors that previously overrode the connector image, command, args, env, volumes, service account, or host namespaces via Spec.PodSpec will see those fields rejected at admission (or silently dropped if the webhook is disabled).
Allowlisted fields flow through unchanged.
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 1.23.0"
},
"package": {
"ecosystem": "Go",
"name": "github.com/fission/fission"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "1.24.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-200",
"CWE-269",
"CWE-284"
],
"github_reviewed": true,
"github_reviewed_at": "2026-06-30T18:12:48Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "### Summary\n\nThe Fission MessageQueueTrigger (MQT) scaler controller exposed two privilege-escalation primitives to any subject able to create MQTs in a namespace.\n\n### Details\n\n**1. Secret materialization.** `getEnvVarlist` in `pkg/mqtrigger/scalermanager.go` read the Secret named in `Spec.Secret` using the controller\u0027s cluster-wide `secrets/get` RBAC and emitted each key as a literal `EnvVar.Value`, copying\nthe plaintext secret content into the connector Deployment\u0027s pod template. A subject holding `messagequeuetriggers/create` but **not** `secrets/get` could exfiltrate any Secret in the namespace by pointing an MQT at it.\n\n**2. PodSpec injection.** `Spec.PodSpec` was merged into the controller-built connector PodSpec via `util.MergePodSpec` with no allowlist on which fields could come from the user. An MQT could substitute `Containers[].Image` (run any\nimage), override `Command`/`Args`, inject `Env`, add `VolumeMounts` + `Volumes`, override `ServiceAccountName`, and set `HostNetwork`/`HostPID`/`HostIPC` \u2014 turning `messagequeuetriggers/create` into effective `deployments/create` with an\n arbitrary image and service account.\n\n### Impact\n\nA tenant with only `messagequeuetriggers.fission.io/create` in a namespace could read any Secret in that namespace and run an arbitrary container image under an arbitrary service account, escalating well beyond their intended RBAC.\n\n### Fix\n\nFixed in [#3367](https://github.com/fission/fission/pull/3367) and released in [v1.24.0](https://github.com/fission/fission/releases/tag/v1.24.0).\n\n- `getEnvVarlist` now emits `EnvVar.ValueFrom.SecretKeyRef` so the connector pod resolves values at start time under its own service account. The secret values are never written into the Deployment object and never logged.\n- A new allowlist, `MergeAllowedPodSpecFields` (`pkg/executor/util/merge_allowlist.go`), accepts only `NodeSelector`, `Tolerations`, `Affinity`, `RuntimeClassName`, and per-container `Resources`. All other user-supplied fields are\ndropped at the controller layer, and the validating webhook rejects every populated non-allowlisted field with a clear error. The webhook and the merge helper share a single canonical `DisallowedPodSpecFields` enumeration so they cannot\ndrift.\n\n### Behavioural change\n\nMQT authors that previously overrode the connector image, command, args, env, volumes, service account, or host namespaces via `Spec.PodSpec` will see those fields rejected at admission (or silently dropped if the webhook is disabled).\nAllowlisted fields flow through unchanged.",
"id": "GHSA-7m8x-qg2j-4m3v",
"modified": "2026-06-30T18:12:48Z",
"published": "2026-06-30T18:12:48Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/fission/fission/security/advisories/GHSA-7m8x-qg2j-4m3v"
},
{
"type": "WEB",
"url": "https://github.com/fission/fission/pull/3367"
},
{
"type": "WEB",
"url": "https://github.com/fission/fission/commit/94bf5792396989fdd71961e8701833c8110da889"
},
{
"type": "PACKAGE",
"url": "https://github.com/fission/fission"
},
{
"type": "WEB",
"url": "https://github.com/fission/fission/releases/tag/v1.24.0"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N",
"type": "CVSS_V3"
}
],
"summary": "Fission: MessageQueueTrigger scaler manager materializes Secret values into Deployment envvars and accepts arbitrary user PodSpec"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.