GHSA-4W7M-58CG-CMFF
Vulnerability from github – Published: 2026-03-13 15:47 – Updated: 2026-03-13 15:47Summary
In affected versions of openclaw, sandboxed leaf subagents could still access the subagents control surface and resolve against the parent requester scope instead of remaining confined to their own session tree.
Impact
A low-privilege sandboxed leaf worker could steer or kill a sibling run owned by the same requester and cause that sibling to execute with its own broader tool policy. This is a sandbox and session-scope boundary bypass.
Affected Packages and Versions
- Package:
openclaw(npm) - Affected versions:
<= 2026.3.8 - Fixed in:
2026.3.11
Technical Details
Leaf subagents retained the subagents tool, and subagent control requests were authorized against the parent requester scope rather than the caller's own spawned descendants. The control path prevented only self-targeting, not cross-sibling steering.
Fix
OpenClaw now removes subagents control access from leaf subagents by default, scopes subagent control to the caller's own descendants, and rejects steer and kill requests that target runs outside that descendant tree. The fix shipped in openclaw@2026.3.11.
Workarounds
Upgrade to 2026.3.11 or later.
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "openclaw"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2026.3.11"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-269"
],
"github_reviewed": true,
"github_reviewed_at": "2026-03-13T15:47:34Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "## Summary\nIn affected versions of `openclaw`, sandboxed leaf subagents could still access the `subagents` control surface and resolve against the parent requester scope instead of remaining confined to their own session tree.\n\n## Impact\nA low-privilege sandboxed leaf worker could steer or kill a sibling run owned by the same requester and cause that sibling to execute with its own broader tool policy. This is a sandbox and session-scope boundary bypass.\n\n## Affected Packages and Versions\n- Package: `openclaw` (npm)\n- Affected versions: `\u003c= 2026.3.8`\n- Fixed in: `2026.3.11`\n\n## Technical Details\nLeaf subagents retained the `subagents` tool, and subagent control requests were authorized against the parent requester scope rather than the caller\u0027s own spawned descendants. The control path prevented only self-targeting, not cross-sibling steering.\n\n## Fix\nOpenClaw now removes `subagents` control access from leaf subagents by default, scopes subagent control to the caller\u0027s own descendants, and rejects `steer` and `kill` requests that target runs outside that descendant tree. The fix shipped in `openclaw@2026.3.11`.\n\n## Workarounds\nUpgrade to `2026.3.11` or later.",
"id": "GHSA-4w7m-58cg-cmff",
"modified": "2026-03-13T15:47:35Z",
"published": "2026-03-13T15:47:34Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/security/advisories/GHSA-4w7m-58cg-cmff"
},
{
"type": "PACKAGE",
"url": "https://github.com/openclaw/openclaw"
},
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/releases/tag/v2026.3.11"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "OpenClaw: Leaf subagents could steer sibling sessions across sandbox boundaries"
}
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.