Vulnerability · Published Aug 21, 2026 · Assigned by GitHub_M

CVE-2026-62675Omnigent: Uploaded Agent Bundle Allows Authenticated Runner RCE via Python Callable Tools

High 8.8

In plain English

A high-severity code injection in omnigent from omnigent-ai. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 0.7% (higher than 51% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.

Severity (CVSS 3.1)8.8High
Exploit odds (30 days)0.7%top 48.7% of all CVEs
Exploited in the wildNo reportsCISA SSVC: poc
Attack toolingNoneno public tooling known

Official description

Omnigent is an open-source AI agent framework and meta-harness for orchestrating coding agents. Prior to 0.3.0, multipart POST /v1/sessions accepts an authenticated user's agent bundle and omnigent/server/bundles.py validate_agent_bundle does not reject a tools..callable dotted Python path. omnigent/runner/tool_dispatch.py _resolve_spec_callable imports the specified module and _execute_spec_callable_tool invokes the resolved function, allowing a bundle to select subprocess.check_output and execute a local command with the runner process permissions. This can expose runner files, environment variables, credentials, workspace data, internal services, and availability without administrator access. This issue is fixed in version 0.3.0.

Affected packages & fixed versions

EcosystemPackageFixed inAdvisory
PyPIomnigent0.3.0GHSA-756x-9hf6-q4h4
PyPIomnigent0.3.0PYSEC-2026-3872

Affected products (as reported)

VendorProductVersions
omnigent-ai omnigent < 0.3.0

Weakness: CWE-94 · Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

References