Every CVE Search/ Get API key

CVE-2025-67510

MySQLWriteTool allows arbitrary/destructive SQL when exposed to untrusted prompts (agent “footgun”)

9.4 CriticalFix available

Neuron is a PHP framework for creating and orchestrating AI Agents. In versions 2.8.11 and below, the MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions. This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions). Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges are impacted. This issue is fixed in version 2.8.12.

Attack profile

How easy it is to exploit

Attack vector
Network
Complexity
Low
Privileges needed
None
User interaction
None

What an attacker can affect

Confidentiality
Low
Integrity
High
Availability
High

Exploit likelihood

0.41%Up 0.3 points since Dec 11, 2025
90%50%10%1%

What each source reports

CVE List Score used 9.4

MySQLWriteTool allows arbitrary/destructive SQL when exposed to untrusted prompts (agent “footgun”)

Neuron is a PHP framework for creating and orchestrating AI Agents. In versions 2.8.11 and below, the MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions. This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions). Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges are impacted. This issue is fixed in version 2.8.12.

Record
CVE-2025-67510
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Weakness
CWE-250, CWE-284
Updated
2026-11-02
EUVDEUVD-2025-202170 9.4

Neuron is a PHP framework for creating and orchestrating AI Agents. In versions 2.8.11 and below, the MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions. This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions). Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges are impacted. This issue is fixed in version 2.8.12.

Record
EUVD-2025-202170
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Updated
2026-11-02
GitHub AdvisoriesGHSA-898v-775g-777c 9.4

Neuron MySQLWriteTool allows arbitrary/destructive SQL when exposed to untrusted prompts (agent “footgun”)

Impact

MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions.

This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions).

Who is impacted: Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges.

Patches

Not patched in: 2.8.11

Recommended improvements (even if keeping the tool intentionally powerful):

  • Provide a safer API that supports only constrained operations (e.g., insertRecord, updateRecord) with allowlisted tables/columns.

  • Add a policy/allowlist layer (e.g., allow only INSERT/UPDATE on selected tables; forbid DROP/TRUNCATE/ALTER/GRANT).

  • Add optional review workflow: log + require human approval for high-risk statements; or “dry-run” mode.

  • Document strongly that the tool must not be exposed to untrusted prompts without additional safeguards.

Workarounds

  • Do not enable MySQLWriteTool for public/untrusted agents.

  • Use a dedicated DB user with least privilege:

    • no DROP, no ALTER, no GRANT, no access to sensitive tables unless necessary
  • Add an application-layer policy rejecting high-risk statements (DROP, TRUNCATE, ALTER, GRANT, REVOKE, CREATE USER, etc.).

  • Implement authorization gating for tool calls (RBAC, allow tool use only for trusted operators).

Record
GHSA-898v-775g-777c
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Weakness
CWE-250, CWE-284
Updated
2026-11-02
NVD 9.4

Neuron is a PHP framework for creating and orchestrating AI Agents. In versions 2.8.11 and below, the MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions. This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions). Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges are impacted. This issue is fixed in version 2.8.12.

Record
CVE-2025-67510
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Weakness
CWE-250, CWE-284
Updated
2026-11-02
OSV 9.4

MySQLWriteTool allows arbitrary/destructive SQL when exposed to untrusted prompts (agent “footgun”)

Neuron is a PHP framework for creating and orchestrating AI Agents. In versions 2.8.11 and below, the MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions. This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions). Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges are impacted. This issue is fixed in version 2.8.12.

Record
CVE-2025-67510
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Weakness
CWE-250, CWE-284
Updated
2026-11-02
OSVGHSA-898v-775g-777c 9.4

Neuron MySQLWriteTool allows arbitrary/destructive SQL when exposed to untrusted prompts (agent “footgun”)

Impact

MySQLWriteTool executes arbitrary SQL provided by the caller using PDO::prepare() + execute() without semantic restrictions.

This is consistent with the name (“write tool”), but in an LLM/agent context it becomes a high-risk capability: prompt injection or indirect prompt manipulation can cause execution of destructive queries such as DROP TABLE, TRUNCATE, DELETE, ALTER, or privilege-related statements (subject to DB permissions).

Who is impacted: Deployments that expose an agent with MySQLWriteTool enabled to untrusted input and/or run the tool with a DB user that has broad privileges.

Patches

Not patched in: 2.8.11

Recommended improvements (even if keeping the tool intentionally powerful):

  • Provide a safer API that supports only constrained operations (e.g., insertRecord, updateRecord) with allowlisted tables/columns.

  • Add a policy/allowlist layer (e.g., allow only INSERT/UPDATE on selected tables; forbid DROP/TRUNCATE/ALTER/GRANT).

  • Add optional review workflow: log + require human approval for high-risk statements; or “dry-run” mode.

  • Document strongly that the tool must not be exposed to untrusted prompts without additional safeguards.

Workarounds

  • Do not enable MySQLWriteTool for public/untrusted agents.

  • Use a dedicated DB user with least privilege:

    • no DROP, no ALTER, no GRANT, no access to sensitive tables unless necessary
  • Add an application-layer policy rejecting high-risk statements (DROP, TRUNCATE, ALTER, GRANT, REVOKE, CREATE USER, etc.).

  • Implement authorization gating for tool calls (RBAC, allow tool use only for trusted operators).

Record
GHSA-898v-775g-777c
CVSS 3.1
9.4 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:H
Weakness
CWE-250, CWE-284
Updated
2026-11-02

Affected products

Affected packages

Affected packages
PackageAffectedFixed
neuron-core/neuron-ai
Packagist
through 2.8.11
2.8.12
neuron-core/neuron-ai
Packagist
before 2.8.12, plus listed versions
1.0.01.0.11.1.01.1.11.1.2
Show 45 moreShow less
1.1.31.10.01.10.11.10.101.10.111.10.121.10.131.10.141.10.151.10.161.10.171.10.181.10.191.10.21.10.201.10.31.10.41.10.51.10.61.10.71.10.81.10.91.11.01.11.11.11.101.11.21.11.31.11.41.11.51.11.61.11.71.11.81.11.91.12.01.12.11.12.101.12.111.12.121.12.131.12.141.12.151.12.161.12.171.12.181.12.2
2.8.12

History

History
RecordedChangedSources
first indexedCVE List, EUVD, GitHub Advisories, NVD, OSV

Dates are when Every CVE recorded each change. Upstream dates may differ.

In short

CVE-2025-67510 is a critical vulnerability (CVSS 9.4) in Neuron-ai Neuron, published December 9, 2025. No source lists public exploit code. EPSS puts the chance of exploitation in the next 30 days at 0.41%, the 35th percentile. Fixed in neuron-core/neuron-ai 2.8.12.

References

This page doesn't list everything

Left out here: 257 versions. The JSON record has them.

MoveEnter Open