GHSA-2FGQ-7J6H-9RM4
Vulnerability from github – Published: 2026-03-03 00:40 – Updated: 2026-03-30 13:19Summary
system.run allowed SHELLOPTS + PS4 environment injection to trigger command substitution during bash -lc xtrace expansion before the allowlisted command body executed.
Affected Packages / Versions
- Package:
openclaw(npm) - Affected:
<= 2026.2.21-2(includes latest published npm version at triage time) - Patched (planned next release):
2026.2.22
Impact
In allowlist mode, an attacker who can invoke system.run with request-scoped env could execute additional shell commands outside the intended allowlisted command body.
Root Cause
Host exec env sanitization blocked startup-file vectors (BASH_ENV, ENV, etc.) but did not block SHELLOPTS/PS4. For shell wrappers (bash|sh|zsh ... -c/-lc), request env overrides were passed through and bash evaluated PS4 under xtrace, enabling command substitution.
Fix
- Block
SHELLOPTSandPS4in host exec env sanitizers (Node + macOS). - For shell wrappers (
bash|sh|zsh ... -c/-lc), reduce request-scoped env overrides to an explicit allowlist (TERM,LANG,LC_*,COLORTERM,NO_COLOR,FORCE_COLOR). - Add regression tests for TS and macOS paths.
Fix Commit(s)
e80c803fa887f9699ad87a9e906ab5c1ff85bd9a
Release Process Note
patched_versions is pre-set to the planned next release (2026.2.22). Once npm release 2026.2.22 is published, advisory publication is a final state action only.
Severity Rationale
This advisory is rated medium because exploitation requires a caller that can already invoke system.run with request-scoped env.
Under OpenClaw's documented trust model (SECURITY.md), authenticated Gateway callers are treated as trusted operators, and adversarial multi-operator / prompt-injection scenarios are out of scope.
The bug remains a real allowlist-intent bypass, but it does not cross a separate trust boundary in the documented deployment assumptions.
OpenClaw thanks @tdjackey for reporting.
{
"affected": [
{
"package": {
"ecosystem": "npm",
"name": "openclaw"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2026.2.22"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-32003"
],
"database_specific": {
"cwe_ids": [
"CWE-15",
"CWE-78"
],
"github_reviewed": true,
"github_reviewed_at": "2026-03-03T00:40:56Z",
"nvd_published_at": "2026-03-19T22:16:32Z",
"severity": "HIGH"
},
"details": "### Summary\n`system.run` allowed `SHELLOPTS` + `PS4` environment injection to trigger command substitution during `bash -lc` xtrace expansion before the allowlisted command body executed.\n\n### Affected Packages / Versions\n- Package: `openclaw` (npm)\n- Affected: `\u003c= 2026.2.21-2` (includes latest published npm version at triage time)\n- Patched (planned next release): `2026.2.22`\n\n### Impact\nIn `allowlist` mode, an attacker who can invoke `system.run` with request-scoped `env` could execute additional shell commands outside the intended allowlisted command body.\n\n### Root Cause\nHost exec env sanitization blocked startup-file vectors (`BASH_ENV`, `ENV`, etc.) but did not block `SHELLOPTS`/`PS4`. For shell wrappers (`bash|sh|zsh ... -c/-lc`), request env overrides were passed through and `bash` evaluated `PS4` under `xtrace`, enabling command substitution.\n\n### Fix\n- Block `SHELLOPTS` and `PS4` in host exec env sanitizers (Node + macOS).\n- For shell wrappers (`bash|sh|zsh ... -c/-lc`), reduce request-scoped env overrides to an explicit allowlist (`TERM`, `LANG`, `LC_*`, `COLORTERM`, `NO_COLOR`, `FORCE_COLOR`).\n- Add regression tests for TS and macOS paths.\n\n### Fix Commit(s)\n- `e80c803fa887f9699ad87a9e906ab5c1ff85bd9a`\n\n### Release Process Note\n`patched_versions` is pre-set to the planned next release (`2026.2.22`). Once npm release `2026.2.22` is published, advisory publication is a final state action only.\n\n### Severity Rationale\nThis advisory is rated **medium** because exploitation requires a caller that can already invoke `system.run` with request-scoped `env`.\n\nUnder OpenClaw\u0027s documented trust model (`SECURITY.md`), authenticated Gateway callers are treated as trusted operators, and adversarial multi-operator / prompt-injection scenarios are out of scope.\n\nThe bug remains a real allowlist-intent bypass, but it does not cross a separate trust boundary in the documented deployment assumptions.\n\nOpenClaw thanks @tdjackey for reporting.",
"id": "GHSA-2fgq-7j6h-9rm4",
"modified": "2026-03-30T13:19:17Z",
"published": "2026-03-03T00:40:56Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/security/advisories/GHSA-2fgq-7j6h-9rm4"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-32003"
},
{
"type": "WEB",
"url": "https://github.com/openclaw/openclaw/commit/e80c803fa887f9699ad87a9e906ab5c1ff85bd9a"
},
{
"type": "PACKAGE",
"url": "https://github.com/openclaw/openclaw"
},
{
"type": "WEB",
"url": "https://www.vulncheck.com/advisories/openclaw-remote-code-execution-via-shellopts-ps4-environment-injection-in-system-run"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:H/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
},
{
"score": "CVSS:4.0/AV:N/AC:H/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
"type": "CVSS_V4"
}
],
"summary": "OpenClaw has system.run shell-wrapper env injection via SHELLOPTS/PS4 can bypass allowlist intent (RCE)"
}
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.