Security readout for executives and security teams
Plain-English summary
CVE-2026-27952 affects self-hosted Agenta API servers before 0.48.1. An authenticated user could break out of Agenta’s Python evaluator sandbox and run code on the API server. This can compromise confidentiality, integrity, and availability of the Agenta server. The SDK used standalone is not reported as affected.
Executive priority
Treat as high priority for any self-hosted Agenta environment with multiple users or externally reachable access. The issue allows authenticated remote code execution on the API server, so remediation should be scheduled urgently, especially where evaluator functionality is enabled.
Technical view
Agenta-API used RestrictedPython for user-supplied evaluator code and allowed numpy inside the sandbox. The advisory states this exposed Python introspection through numpy.ma.core.inspect, enabling access to system-level functionality and arbitrary code execution in the API process. Fixed in 0.48.1 by removing numpy from the sandbox allowlist; later 0.60+ versions use a different execution model.
Likely exposure
Exposure is limited to Agenta self-hosted platform deployments running agenta-api before 0.48.1 where authenticated users can submit or manage custom evaluator code. Standalone SDK usage is not identified as affected in the provided sources.
Exploitation context
The CVSS vector indicates network access, low complexity, authenticated privileges required, and no user interaction. The provided sources do not report active exploitation, and the CVE is not marked as CISA KEV in the bundle.
Researcher notes
The issue is CWE-94 code injection in a server-side custom evaluator. Analysis should focus on affected agenta-api versions before 0.48.1 and access paths to custom evaluator execution. Do not assume cloud, SDK, or unrelated Agenta components are affected without vendor confirmation.
Mitigation direction
- Upgrade agenta-api to version 0.48.1 or later.
- Consider 0.60+ where the older RestrictedPython sandbox model was removed.
- Restrict evaluator-code access to trusted users until upgraded.
- Review vendor advisory for deployment-specific guidance.
Validation and detection
- Inventory self-hosted Agenta API server versions.
- Confirm whether agenta-api is below 0.48.1.
- Identify users allowed to create or run custom evaluator code.
- Review API server logs for unusual evaluator activity.
- Verify upgrade by checking the deployed agenta-api version.
Public sources used
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
CWE-94: Code execution behavior lookup
Code execution and unsafe deserialization weaknesses often justify reviewing execution behavior and process telemetry. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupExecution behavior lookup
The CVE wording references code or command execution, so execution technique review may help defensive triage. This is a Glexia inferred lookup path, not an official MITRE, ATT&CK, or CVE Program mapping.
Open ATT&CK lookupPrivilege behavior lookup
The CVE wording references privilege impact, so privilege escalation and authorization behavior review may help. This is a Glexia inferred lookup path, not an official MITRE, ATT&CK, or CVE Program mapping.
Open ATT&CK lookupCVE-2026-27952 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
Open ATT&CK lookup- Severity
- High
- CVSS
- 8.8 (3.1)
- Known Exploited
- No
- Published
Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
CNA and ADP enrichment extracted from CVE v5
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
CVSS vector scores
1 official scoreWe collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H2.85.9Primary CVE scoreVulnerability scoring details
Base CVSS 3.1 score
8.8HighVector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Source materials
- CVE List V5 sourceCVE List V5
- https://github.com/Agenta-AI/agenta/security/advisories/GHSA-pmgp-2m3v-34mqCVE reference · x_refsource_CONFIRM
Products and packages named in the record
CWE details
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.
Improper Control of Generation of Code ('Code Injection')
Improper Control of Generation of Code ('Code Injection') represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.