GCVE-1988-2026-0075
Vulnerability from gna-1988 – Published: 2026-09-07 13:20 – Updated: 2026-09-07 13:20| Vendor | Product | Version | CPE status | |
|---|---|---|---|---|
| unknown | Chronicle Wire |
Affected:
unknown
|
guessed |
{
"containers": {
"cna": {
"affected": [
{
"product": "Chronicle Wire",
"vendor": "unknown",
"versions": [
{
"status": "affected",
"version": "unknown"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "Ron E"
}
],
"descriptions": [
{
"lang": "en",
"value": "Chronicle Wire permits YAML type tags supplied within serialized input to\ninfluence Java class selection and object instantiation during untyped\ndeserialization.\n\nWhen applications deserialize attacker-controlled or otherwise untrusted\nYAML through APIs such as readObject() or object(Object.class), an\nexternally controlled YAML type tag can identify a Java class that\nChronicle Wire resolves through its configured ClassLookup.\n\nWhen the default permissive class lookup is used and the supplied class can\nbe resolved, the resulting class can propagate through Chronicle Wire\u0027s\ngeneric object deserialization path and ultimately reach\nObjectUtils.newInstance(clazz). Chronicle Wire can then invoke the selected\nclass\u0027s deserialization lifecycle, including readMarshallable() where\napplicable.\n\nThe security-sensitive behavior is therefore not limited to ordinary data\nbinding. Externally supplied serialized data can influence *which Java\nclass is instantiated during deserialization*.\n\nThe accompanying proof of concept confirms that:\n\n - An externally controlled YAML type tag selects the Java class\n instantiated by readObject().\n - The selected class\u0027s constructor is executed.\n - A selected class\u0027s readMarshallable() implementation is automatically\n invoked during deserialization.\n - An existing third-party class already present on the runtime classpath\n can be instantiated using a YAML type tag.\n - Configuring a restrictive ClassLookup prevents the demonstrated\n arbitrary class selection.\n\nThe practical security impact depends on the classes available on the\ntarget application\u0027s classpath and whether untrusted YAML reaches an\naffected untyped deserialization API. The PoC establishes the arbitrary\nclass-selection and instantiation primitive but does not claim universal\narbitrary code execution.\nVulnerability Details\n\nChronicle Wire supports YAML type tags capable of identifying Java classes\nduring deserialization.\n\nFor example:\n\n!fully.qualified.ClassName\n\nThe parser does not treat this value solely as descriptive metadata. The\nsupplied class name is resolved using the wire\u0027s configured ClassLookup.\n\nThe default wire configuration initializes the lookup using the global\nalias pool:\n\nprotected ClassLookup classLookup =\n ClassAliasPool.CLASS_ALIASES;\n\nWhen a YAML TAG token is encountered, the supplied type is resolved:\n\nClass\u003c?\u003e typePrefix() {\n ...\n return classLookup().forName(stringBuilder);\n}\n\nThe class represented by the serialized YAML can therefore influence the\nJava type selected during deserialization.\n\nIn affected object-reading paths, Chronicle Wire can subsequently\ninstantiate the resolved class:\n\nClass\u003c?\u003e clazz = typePrefix();\n\nif (clazz != object.getClass())\n object = ObjectUtils.newInstance(clazz);\n\nThe externally selected type can also propagate into the generic object\ndeserialization path:\n\nObject o = typePrefixOrObject(clazz);\n\n...\n\nt = Wires.object2(..., (Class) o);\n\nWithin Wires.object2(), the type supplied by serialized input can replace\nthe caller\u0027s original type under several conditions:\n\nif (clazz == null\n || clazz.isAssignableFrom(clazz2)\n || ReadResolvable.class.isAssignableFrom(clazz2)\n || !ObjectUtils.isConcreteClass(clazz))\n{\n clazz = clazz2;\n}\n\nChronicle Wire can then instantiate the selected class:\n\nif (o == null)\n o = ObjectUtils.newInstance(clazz);\n\nand continue the object\u0027s deserialization lifecycle:\n\nWires.readMarshallable(\n clazz,\n o,\n in.wireIn(),\n true);\n\nAs a result, when permissive class resolution is available, externally\ncontrolled YAML can influence both the class instantiated by Chronicle Wire\nand the class-specific deserialization logic subsequently executed.\nRoot Cause\n\nThe root cause is the use of serialized YAML type information to select\nJava classes during generic or untyped object deserialization without a\nmandatory deny-by-default class allow-list.\n\nChronicle Wire resolves externally supplied YAML type tags through its\nconfigured ClassLookup. When the default permissive lookup permits the\nrequested type, the resulting class can propagate into generic object\ndeserialization and reach ObjectUtils.newInstance().\n\nThe security boundary becomes particularly important when an application\nperforms operations such as:\n\nTextWire.from(untrustedYaml).readObject();\n\nor equivalent untyped deserialization.\n\nIn this situation, the application is not exclusively determining the Java\nclass being constructed. The serialized YAML participates in that decision.\n\nA restrictive ClassLookup can prevent arbitrary class resolution, but such\na restriction is not inherent to the demonstrated default deserialization\npath.\nProof of Concept Results\n\nThe security test suite successfully reproduced multiple independent paths\nin which serialized type information caused Chronicle Wire to instantiate\nclasses selected through the supplied Wire/YAML data.\n\nThe tests completed successfully with no failures or errors:\n\n[INFO] Running net.openhft.chronicle.wire.SecurityAdditionalPoCTest\n\n[INFO] Tests run: 13, Failures: 0, Errors: 0, Skipped: 0\n\n[INFO] BUILD SUCCESS\n\nDirect Tagged Class Instantiation\n\nThe PoC confirmed that Chronicle Wire resolves a supplied YAML type tag and\ninstantiates the corresponding Java class during deserialization.\n\nObserved result:\n\nWireObjectInput.readObject instantiated tagged class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$AdditionalTypedPathProbe\n\nThis confirms that the class encoded in serialized input is not merely\nparsed as metadata. The resolved type reaches object construction and\nresults in an instance of the tagged class.\nMap Value Type Instantiation\n\nThe same externally controlled type-selection behavior was reproduced while\ndeserializing a typed value contained within a map.\n\nObserved result:\n\nMap value read instantiated tagged class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$AdditionalTypedPathProbe\n\nThis demonstrates that the behavior is not limited to a single top-level\nreadObject() operation. Typed serialized values encountered within other\nobject-reading paths can also cause tagged classes to be instantiated.\nFile-Based Typed Deserialization\n\nThe PoC additionally confirmed arbitrary class instantiation when typed\nserialized data is loaded from a caller-controlled file.\n\nThe test file contained:\n\n!net.openhft.chronicle.wire.SecurityAdditionalPoCTest$FilePathProbe {\n marker: from-file\n}\n\nThe test output confirmed the exact payload written to the file:\n\nWireType.fromFile payload written BEGIN\n!net.openhft.chronicle.wire.SecurityAdditionalPoCTest$FilePathProbe {\nmarker: from-file }\nWireType.fromFile payload written END\n\nChronicle Wire subsequently instantiated the class identified by the\nserialized type tag:\n\nWireType.fromFile instantiated tagged class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$FilePathProbe\n\nThis provides an additional concrete deserialization path where serialized\ntype information determines the Java class instantiated by Chronicle Wire.\nStream-Based File Deserialization\n\nThe same behavior was confirmed through the file-stream deserialization\npath.\n\nObserved result:\n\nWireType.streamFromFile instantiated tagged class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$AdditionalTypedPathProbe\n\nThis demonstrates that externally supplied type information can reach class\ninstantiation through more than one Chronicle Wire input API.\nConstructor and readMarshallable() Execution\n\nThe strongest lifecycle evidence was produced through a MethodReader typed\nargument.\n\nThe serialized argument selected the following class:\n\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$MethodReaderProbe\n\nThe test recorded both constructor and deserialization callback invocation:\n\nMethodReader typed argument instantiated class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$MethodReaderProbe\n\nMethodReader typed argument constructor calls: 1\nMethodReader typed argument readMarshallable calls: 1\n\nThis confirms that externally selected type information can result in more\nthan creation of an inert Java object.\n\nFor the selected class, Chronicle Wire caused:\n\n -\n\n the class to be resolved;\n -\n\n an instance to be constructed;\n -\n\n the constructor to execute; and\n -\n\n the class-specific readMarshallable() callback to execute.\n\nThe observed invocation counts were:\n\nConstructor calls: 1\nreadMarshallable calls: 1\n\nThis provides direct evidence that class-specific executable lifecycle\nbehavior is reached as a consequence of serialized type selection.\nConsolidated PoC Evidence\n\nThe test results demonstrate arbitrary class instantiation through several\nChronicle Wire deserialization paths:\n\n -\n\n WireObjectInput.readObject() instantiated an externally tagged class.\n -\n\n Map value deserialization instantiated an externally tagged class.\n -\n\n WireType.fromFile() instantiated the class identified by a YAML type tag\n contained in a caller-controlled file.\n -\n\n WireType.streamFromFile() instantiated an externally tagged class.\n -\n\n MethodReader typed argument deserialization instantiated an externally\n selected class.\n -\n\n Constructor execution was directly observed.\n -\n\n readMarshallable() execution was directly observed.\n -\n\n All 13 security tests completed without failure or error.\n\nThe most significant lifecycle result was:\n\nMethodReader typed argument instantiated class:\nnet.openhft.chronicle.wire.SecurityAdditionalPoCTest$MethodReaderProbe\n\nMethodReader typed argument constructor calls: 1\nMethodReader typed argument readMarshallable calls: 1\n\nCombined with the direct readObject(), map-value, fromFile(), and\nstreamFromFile() results, the PoC demonstrates that externally supplied\ntype information can influence Java class selection and cause the selected\nclass to be instantiated across multiple Chronicle Wire deserialization\nsurfaces.\n\nThe PoC does not rely solely on inspecting the source code or confirming\nthat a class name was successfully resolved. It observes actual object\nconstruction and class-specific deserialization callback execution at\nruntime.\nPoC Conclusion\n\nThe test suite confirms the core security primitive described by this\nfinding: *serialized type information can control which Java class\nChronicle Wire instantiates during affected deserialization operations*.\n\nThe runtime evidence confirms both object construction and execution of\nclass-specific deserialization lifecycle behavior:\n\n[CONFIRMED] Externally supplied type selected Java class\n[CONFIRMED] Selected class instantiated\n[CONFIRMED] Constructor executed\n[CONFIRMED] readMarshallable() executed\n[CONFIRMED] Typed class instantiated through readObject()\n[CONFIRMED] Typed class instantiated through map value deserialization\n[CONFIRMED] Typed class instantiated through WireType.fromFile()\n[CONFIRMED] Typed class instantiated through WireType.streamFromFile()\n[CONFIRMED] 13 security tests completed with 0 failures and 0 errors\n\nThese results establish the *arbitrary class selection and instantiation\nprimitive*. The ultimate security impact remains dependent on the classes\navailable on the target application\u0027s classpath and the trust boundary\nthrough which serialized input reaches Chronicle Wire.\n\nRon Edgerson\nVulnerability Researcher \u0026 Exploit Developer\n\nCVE Research | Binary Exploitation | Application \u0026 Systems Security\nResponsible Disclosure \u2022 Proof-of-Concept Development\n\n\ud83c\udf10 https://github.com/ob1sec\n\ud83d\udd17 https://www.linkedin.com/in/ronedgerson1\n\u003chttps://linkedin.com/in/yourhandle\u003e\n_______________________________________________\nSent through the Full Disclosure mailing list\nhttps://nmap.org/mailman/listinfo/fulldisclosure\nWeb Archives \u0026 RSS: https://seclists.org/fulldisclosure/"
}
],
"providerMetadata": {
"dateUpdated": "2026-09-07T13:20:21Z",
"orgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"shortName": "VULNARCHIVE"
},
"references": [
{
"tags": [
"technical-description",
"exploit"
],
"url": "https://vuln.freearchive.org/archive/full-disclosure/2026/Aug/103"
},
{
"tags": [
"technical-description"
],
"url": "https://seclists.org/fulldisclosure/2026/Aug/103"
},
{
"url": "https://github.com/ob1sec"
},
{
"url": "https://linkedin.com/in/yourhandle"
},
{
"url": "https://nmap.org/mailman/listinfo/fulldisclosure"
},
{
"url": "https://seclists.org/fulldisclosure/"
},
{
"url": "https://www.linkedin.com/in/ronedgerson1"
}
],
"source": {
"defect": [
"https://seclists.org/fulldisclosure/2026/Aug/103"
],
"discovery": "EXTERNAL"
},
"title": "Chronicle Wire v2026.8 Arbitrary Class Instantiation During YAML Deserialization via Externally Controlled YAML Type Tags",
"x_gcve": [
{
"recordType": "advisory",
"relationships": [],
"vulnId": "GCVE-1988-2026-0075",
"x_vulnarchive": {
"archiveUrl": "https://vuln.freearchive.org/archive/full-disclosure/2026/Aug/103",
"automated": true,
"contentSha256": "373857b24efd6a689043e6cf615e0c3a0a62392da12986dc20b6425ac929092f",
"evidenceScore": 9,
"messageId": "",
"originalUrl": "https://seclists.org/fulldisclosure/2026/Aug/103",
"policy": "vulnarchive-1",
"sourceFormat": "text/html",
"sourcePublishedAt": "2026-08-22T12:42:32Z"
}
}
]
}
},
"cveMetadata": {
"assignerOrgId": "4e2abfbf-4a2a-4b76-a4e0-d77c18ba156c",
"assignerShortName": "VULNARCHIVE",
"datePublished": "2026-09-07T13:20:21Z",
"dateUpdated": "2026-09-07T13:20:21Z",
"state": "PUBLISHED",
"vulnId": "GCVE-1988-2026-0075"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
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.