Описание
LangChain4j: SQL injection via metadata filters in langchain4j-mariadb and langchain4j-pgvector
Summary
The MariaDB and pgvector embedding stores build metadata-filter SQL by string-concatenating
filter keys (and, in MariaDB, string values) directly into the query without adequate
escaping. A crafted metadata key in EmbeddingSearchRequest.filter() can break out of its SQL
context and inject arbitrary SQL into the statements executed by the stores' search and
removeAll(Filter) operations.
Details
pgvector — JSON mode (default, COMBINED_JSON / COMBINED_JSONB). JSONFilterMapper
places the key inside a single-quoted SQL literal (the JSON key of the ->> operator) with no
escaping:
A key containing a single quote breaks out, e.g.
metadataKey("')::text IS NOT NULL OR pg_sleep(1) IS NOT NULL --") injects a live pg_sleep(1)
(observable as a delay; exploitable for blind data extraction).
pgvector — column mode (COLUMN_PER_KEY). ColumnFilterMapper used the key as a bare,
unquoted, unvalidated SQL identifier (<key>::<type>), so a key such as 1=1 OR true --
injects directly.
MariaDB — JSON mode (default). JSONFilterMapper placed the key inside the JSON path literal
'$.<key>' unescaped (same break-out mechanism). Additionally, MariaDbFilterMapper.formatValue()
escaped ' but not \; because MariaDB treats backslash as an escape character by default, a
string value ending in a backslash could also break out of its literal.
MariaDB — column mode (COLUMN_PER_KEY). ColumnFilterMapper fell back to the raw,
unescaped key when the driver could not quote it as an identifier (e.g. a
character).
The filter key is the runtime injection surface; both stores' search() (including pgvector's
HYBRID mode) and removeAll(Filter) are affected. Add/upsert operations a
parameterized and not affected.
Impact
Applications that allow attacker-influenced metadata filter keys (e.g. use LLM-generated filters) to reach these stores are exposed to SQL injection: blind data exfiltration, denial of service via sleep functions, and — through `remove deletion of arbitrary rows. Applications using only hard-coded, developer-defined filter keys are not reachable.
Patches
Fixed in langchain4j-mariadb and langchain4j-pgvector 1.16.3-beta26:
- JSON filter keys are escaped before being embedded in the SQL string lit
quotes doubled, correct for PostgreSQL
standard_conforming_strings = on; MariaDB: backslash and single quote). - MariaDB string values escape both
\and'. - Column-mode keys are validated/quoted as identifiers and rejected when u concatenated as raw SQL.
Workarounds
- Do not pass untrusted input as metadata filter keys.
- Restrict filter keys to a known allow-list at the application layer.
References
- pgvector:
JSONFilterMapper,ColumnFilterMapper - MariaDB:
JSONFilterMapper,MariaDbFilterMapper,ColumnFilterMapper
Пакеты
dev.langchain4j:langchain4j-mariadb
<= 1.2.0-beta8
1.2.1-beta8
dev.langchain4j:langchain4j-mariadb
>= 1.3.0-beta9, <= 1.5.0-beta11
1.5.1-beta11
dev.langchain4j:langchain4j-mariadb
>= 1.6.0-beta12, <= 1.11.7-beta19
1.11.8-beta19
dev.langchain4j:langchain4j-mariadb
>= 1.12.1-beta21, <= 1.16.2-beta26
1.16.3-beta26
dev.langchain4j:langchain4j-pgvector
<= 1.2.0-beta8
1.2.1-beta8
dev.langchain4j:langchain4j-pgvector
>= 1.3.0-beta9, <= 1.5.0-beta11
1.5.1-beta11
dev.langchain4j:langchain4j-pgvector
>= 1.6.0-beta12, <= 1.11.7-beta19
1.11.8-beta19
dev.langchain4j:langchain4j-pgvector
>= 1.12.1-beta21, <= 1.16.2-beta26
1.16.3-beta26
Связанные уязвимости
LangChain4j is a Java library for building LLM-powered applications on the JVM. Prior to 1.2.1-beta8, 1.5.1-beta11, 1.11.8-beta19, and 1.16.3-beta26, the MariaDB and pgvector embedding stores build metadata-filter SQL by string-concatenating filter keys, and in MariaDB string values, directly into the query without adequate escaping. A crafted metadata key in EmbeddingSearchRequest.filter() can break out of its SQL context and inject arbitrary SQL into the statements executed by the stores' search and removeAll(Filter) operations, enabling blind data exfiltration, denial of service via sleep functions, and deletion of arbitrary rows through removeAll(Filter). This issue is fixed in langchain4j-mariadb and langchain4j-pgvector versions 1.2.1-beta8, 1.5.1-beta11, 1.11.8-beta19, and 1.16.3-beta26.