CVE-2026-58449

Received Received - Intake

Remote Code Execution in txtai via Reindex API

Publication date: 2026-06-30

Last updated on: 2026-06-30

Assigner: VulnCheck

Description

txtai through 9.10.0, fixed in commit 11b32da, exposes an API /reindex endpoint whose function body parameter is resolved through txtai.util.Resolver, which performs __import__ and getattr on the caller-supplied dotted path with no allowlist. When the API is exposed with no TOKEN configured (authentication is opt-in, so all endpoints are unauthenticated) and the index is configured writable, a remote attacker can set function to an arbitrary callable such as subprocess.getoutput, achieving remote code execution as the server process during reindexing. Exploitation requires those deployment conditions (API exposed, no TOKEN, writable index); it is not the default configuration. The fix gates the endpoint behind a new reindex configuration flag.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published

2026-06-30

Last Modified

2026-06-30

Generated

2026-07-01

AI Q&A

2026-07-01

EPSS Evaluated

N/A

NVD

CVE-2026-58449

Affected Vendors & Products

Vendor	Product	Version / Range
txtai	txtai	to 9.10.0 (exc)

Helpful Resources

Exploitability

CWE

KEV

CWE ID	Description
CWE-94	The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.

Attack-Flow Graph

Executive Summary

This vulnerability exists in txtai versions through 9.10.0 and involves the /reindex API endpoint. The endpoint accepts a function body parameter that is resolved using txtai.util.Resolver, which uses Python's __import__ and getattr functions on a caller-supplied dotted path without any allowlist. If the API is exposed without a TOKEN configured (authentication is optional and off by default) and the index is writable, a remote attacker can specify an arbitrary callable function such as subprocess.getoutput. This allows the attacker to execute arbitrary code remotely on the server process during reindexing.

Exploitation requires specific deployment conditions: the API must be exposed, no authentication token must be set, and the index must be writable. These are not default settings. The vulnerability was fixed by gating the endpoint behind a new reindex configuration flag.

Impact Analysis

This vulnerability can lead to remote code execution on the server running txtai. An attacker who exploits this can run arbitrary commands with the same privileges as the server process, potentially leading to full system compromise, data theft, data manipulation, or disruption of services.

Mitigation Strategies

To mitigate this vulnerability, ensure that the /reindex API endpoint is not exposed without authentication by configuring a TOKEN, as authentication is opt-in and endpoints are unauthenticated by default.

Additionally, avoid configuring the index as writable if not necessary, since a writable index combined with an exposed unauthenticated API allows remote code execution.

Applying the fix that gates the /reindex endpoint behind a new reindex configuration flag is also recommended.

Hi! I’m here to help you understand CVE-2026-58449. Ask me anything about the vulnerability, its impact, or mitigation strategies.

0/70